Usually doesn't need to be instantiated externally. Access QCustomPlot's default four axes via QCustomPlot::xAxis (bottom), QCustomPlot::yAxis (left), QCustomPlot::xAxis2 (top) and QCustomPlot::yAxis2 (right).
Axes are always part of an axis rect, see QCPAxisRect.
Naming convention of axis parts
Overview of the spacings and paddings that define the geometry of an axis. The dashed gray line on the left represents the QCustomPlot widget border.
When automatic tick label generation is enabled (setAutoTickLabels), defines how the coordinate of the tick is interpreted, i.e. translated into a string.
Tick coordinate is regarded as normal number and will be displayed as such. (see setNumberFormat)
ltDateTime
Tick coordinate is regarded as a date/time (seconds since 1970-01-01T00:00:00 UTC) and will be displayed and formatted as such. (for details, see setDateTimeFormat)
Sets whether the axis uses a linear scale or a logarithmic scale. If type is set to stLogarithmic, the logarithm base can be set with setScaleLogBase. In logarithmic axis scaling, major tick marks appear at all powers of the logarithm base. Properties like tick step (setTickStep) don't apply in logarithmic scaling. If you wish a decimal base but less major ticks, consider choosing a logarithm base of 100, 1000 or even higher.
If type is stLogarithmic and the number format (setNumberFormat) uses the 'b' option (beautifully typeset decimal powers), the display usually is "1 [multiplication sign] 10
[superscript] n", which looks unnatural for logarithmic scaling (the "1 [multiplication sign]" part). To only display the decimal power, set the number precision to zero with setNumberPrecision.
void QCPAxis::setScaleLogBase
(
double
base
)
If setScaleType is set to stLogarithmic, base will be the logarithm base of the scaling. In logarithmic axis scaling, major tick marks appear at all powers of base.
Properties like tick step (setTickStep) don't apply in logarithmic scaling. If you wish a decimal base but less major ticks, consider choosing base 100, 1000 or even higher.
This slot may be connected with the rangeChanged signal of another axis so this axis is always synchronized with the other axis range, when it changes.
The position coordinate indicates together with the alignment parameter, where the new range will be positioned. size defines the size of the new axis range. alignment may be Qt::AlignLeft, Qt::AlignRight or Qt::AlignCenter. This will cause the left border, right border, or center of the range to be aligned with position. Any other values of alignment will default to Qt::AlignCenter.
void QCPAxis::setRangeLower
(
double
lower
)
Sets the lower bound of the axis range. The upper bound is not changed.
Sets whether the axis range (direction) is displayed reversed. Normally, the values on horizontal axes increase left to right, on vertical axes bottom to top. When reversed is set to true, the direction of increasing values is inverted.
Note that the range and data interface stays the same for reversed axes, e.g. the lower part of the setRange interface will still reference the mathematically smaller number than the upper part.
void QCPAxis::setAutoTicks
(
bool
on
)
Sets whether the tick positions should be calculated automatically (either from an automatically generated tick step or a tick step provided manually via setTickStep, see setAutoTickStep).
If on is set to false, you must provide the tick positions manually via setTickVector. For these manual ticks you may let QCPAxis generate the appropriate labels automatically by leaving setAutoTickLabels set to true. If you also wish to control the displayed labels manually, set setAutoTickLabels to false and provide the label strings with setTickVectorLabels.
If you need dynamically calculated tick vectors (and possibly tick label vectors), set the vectors in a slot connected to the ticksRequest signal.
void QCPAxis::setAutoTickCount
(
int
approximateCount
)
When setAutoTickStep is true, approximateCount determines how many ticks should be generated in the visible range, approximately.
It's not guaranteed that this number of ticks is met exactly, but approximately within a tolerance of about two.
Only values greater than zero are accepted as approximateCount.
void QCPAxis::setAutoTickLabels
(
bool
on
)
Sets whether the tick labels are generated automatically. Depending on the tick label type (ltNumber or ltDateTime), the labels will either show the coordinate as floating point number (setNumberFormat), or a date/time formatted according to setDateTimeFormat.
If on is set to false, you should provide the tick labels via setTickVectorLabels. This is usually used in a combination with setAutoTicks set to false for complete control over tick positions and labels, e.g. when the ticks should be at multiples of pi and show "2pi", "3pi" etc. as tick labels.
If you need dynamically calculated tick vectors (and possibly tick label vectors), set the vectors in a slot connected to the ticksRequest signal.
void QCPAxis::setAutoTickStep
(
bool
on
)
Sets whether the tick step, i.e. the interval between two (major) ticks, is calculated automatically. If on is set to true, the axis finds a tick step that is reasonable for human readable plots.
The number of ticks the algorithm aims for within the visible range can be set with setAutoTickCount.
If on is set to false, you may set the tick step manually with setTickStep.
void QCPAxis::setAutoSubTicks
(
bool
on
)
Sets whether the number of sub ticks in one tick interval is determined automatically. This works, as long as the tick step mantissa is a multiple of 0.5. When setAutoTickStep is enabled, this is always the case.
When on is set to false, you may set the sub tick count with setSubTickCount manually.
void QCPAxis::setTicks
(
bool
show
)
Sets whether tick marks are displayed.
Note that setting show to false does not imply that tick labels are invisible, too. To achieve that, see setTickLabels.
void QCPAxis::setTickLabels
(
bool
show
)
Sets whether tick labels are displayed. Tick labels are the numbers drawn next to tick marks.
void QCPAxis::setTickLabelPadding
(
int
padding
)
Sets the distance between the axis base line (including any outward ticks) and the tick labels.
In QCustomPlot, date/time coordinates are double numbers representing the seconds since 1970-01-01T00:00:00 UTC. This format can be retrieved from QDateTime objects with the QDateTime::toTime_t() function. Since this only gives a resolution of one second, there is also the QDateTime::toMSecsSinceEpoch() function which returns the timespan described above in milliseconds. Divide its return value by 1000.0 to get a value with the format needed for date/time plotting, with a resolution of one millisecond.
Using the toMSecsSinceEpoch function allows dates that go back to 2nd January 4713 B.C. (represented by a negative number), unlike the toTime_t function, which works with unsigned integers and thus only goes back to 1st January 1970. So both for range and accuracy, use of toMSecsSinceEpoch()/1000.0 should be preferred as key coordinate for date/time axes.
Sets the rotation of the tick labels. If degrees is zero, the labels are drawn normally. Else, the tick labels are drawn rotated by degrees clockwise. The specified angle is bound to values from -90 to 90 degrees.
If degrees is exactly -90, 0 or 90, the tick labels are centered on the tick coordinate. For other angles, the label is drawn with an offset such that it seems to point toward or away from the tick mark.
void QCPAxis::setDateTimeFormat
(
const QString &
format
)
Sets the format in which dates and times are displayed as tick labels, if setTickLabelType is ltDateTime. for details about the format string, see the documentation of QDateTime::toString().
The default value of QDateTime objects (and also QCustomPlot) is Qt::LocalTime. However, if the date time values passed to QCustomPlot are given in the UTC spec, set timeSpec to Qt::UTC to get the correct axis labels.
Sets the number format for the numbers drawn as tick labels (if tick label type is ltNumber). This formatCode is an extended version of the format code used e.g. by QString::number() and QLocale::toString(). For reference about that, see the "Argument Formats" section in the detailed description of the QString class. formatCode is a string of one, two or three characters. The first character is identical to the normal format code used by Qt. In short, this means: 'e'/'E' scientific format, 'f' fixed format, 'g'/'G' scientific or fixed, whichever is shorter.
The second and third characters are optional and specific to QCustomPlot:
If the first char was 'e' or 'g', numbers are/might be displayed in the scientific format, e.g. "5.5e9", which is ugly in a plot. So when the second char of formatCode is set to 'b' (for "beautiful"), those exponential numbers are formatted in a more natural way, i.e. "5.5
[multiplication sign] 10 [superscript] 9". By default, the multiplication sign is a centered dot. If instead a cross should be shown (as is usual in the USA), the third char of formatCode can be set to 'c'. The inserted multiplication signs are the UTF-8 characters 215 (0xD7) for the cross and 183 (0xB7) for the dot.
If the scale type (setScaleType) is stLogarithmic and the formatCode uses the 'b' option (beautifully typeset decimal powers), the display usually is "1 [multiplication sign] 10
[superscript] n", which looks unnatural for logarithmic scaling (the "1 [multiplication sign]" part). To only display the decimal power, set the number precision to zero with setNumberPrecision.
Examples for formatCode:
g normal format code behaviour. If number is small, fixed format is used, if number is large, normal scientific format is used
gb If number is small, fixed format is used, if number is large, scientific format is used with beautifully typeset decimal powers and a dot as multiplication sign
ebc All numbers are in scientific format with beautifully typeset decimal power and a cross as multiplication sign
fb illegal format code, since fixed format doesn't support (or need) beautifully typeset decimal powers. Format code will be reduced to 'f'.
hello illegal format code, since first char is not 'e', 'E', 'f', 'g' or 'G'. Current format code will not be changed.
void QCPAxis::setNumberPrecision
(
int
precision
)
Sets the precision of the tick label numbers. See QLocale::toString(double i, char f, int prec) for details. The effect of precisions are most notably for number Formats starting with 'e', see setNumberFormat
If the scale type (setScaleType) is stLogarithmic and the number format (setNumberFormat) uses the 'b' format code (beautifully typeset decimal powers), the display usually is "1 [multiplication sign] 10 [superscript] n", which looks unnatural for logarithmic scaling (the redundant "1 [multiplication sign]" part). To only display the decimal power "10
[superscript] n", set precision to zero.
void QCPAxis::setTickStep
(
double
step
)
If setAutoTickStep is set to false, use this function to set the tick step manually. The tick step is the interval between (major) ticks, in plot coordinates.
If you want full control over what ticks (and possibly labels) the axes show, this function is used to set the coordinates at which ticks will appear.setAutoTicks must be disabled, else the provided tick vector will be overwritten with automatically generated tick coordinates upon replot. The labels of the ticks can be generated automatically when setAutoTickLabels is left enabled. If it is disabled, you can set the labels manually with setTickVectorLabels.
vec is a vector containing the positions of the ticks, in plot coordinates.
Warning
vec must be sorted in ascending order, no additional checks are made to ensure this.
If you want full control over what ticks and labels the axes show, this function is used to set a number of QStrings that will be displayed at the tick positions which you need to provide with setTickVector. These two vectors should have the same size. (Note that you need to disable setAutoTicks and setAutoTickLabels first.)
vec is a vector containing the labels of the ticks. The entries correspond to the respective indices in the tick vector, passed via setTickVector.
Sets the length of the ticks in pixels. inside is the length the ticks will reach inside the plot and outside is the length they will reach outside the plot. If outside is greater than zero, the tick labels and axis label will increase their distance to the axis accordingly, so they won't collide with the ticks.
Sets the length of the outward ticks in pixels. outside is the length the ticks will reach outside the plot. If outside is greater than zero, the tick labels and axis label will increase their distance to the axis accordingly, so they won't collide with the ticks.
Sets the number of sub ticks in one (major) tick step. A sub tick count of three for example, divides the tick intervals in four sub intervals.
By default, the number of sub ticks is chosen automatically in a reasonable manner as long as the mantissa of the tick step is a multiple of 0.5. When setAutoTickStep is enabled, this is always the case.
If you want to disable automatic sub tick count and use this function to set the count manually, see setAutoSubTicks.
void QCPAxis::setSubTickLength
(
int
inside,
int
outside = 0
)
Sets the length of the subticks in pixels. inside is the length the subticks will reach inside the plot and outside is the length they will reach outside the plot. If outside is greater than zero, the tick labels and axis label will increase their distance to the axis accordingly, so they won't collide with the ticks.
void QCPAxis::setSubTickLengthIn
(
int
inside
)
Sets the length of the inward subticks in pixels. inside is the length the subticks will reach inside the plot.
Sets the length of the outward subticks in pixels. outside is the length the subticks will reach outside the plot. If outside is greater than zero, the tick labels will increase their distance to the axis accordingly, so they won't collide with the ticks.
Sets the text of the axis label that will be shown below/above or next to the axis, depending on its orientation. To disable axis labels, pass an empty string as str.
void QCPAxis::setLabelPadding
(
int
padding
)
Sets the distance between the tick labels and the axis label.
Sets the offset the axis has to its axis rect side.
If an axis rect side has multiple axes, only the offset of the inner most axis has meaning. The offset of the other axes is controlled automatically, to place the axes at appropriate positions to prevent them from overlapping.
void QCPAxis::setSelectedTickLabelFont
(
const QFont &
font
)
Sets the font that is used for tick labels when they are selected.
Sets whether the user can (de-)select the parts in selectable by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains iSelectAxes.)
However, even when selectable is set to a value not allowing the selection of a specific part, it is still possible to set the selection of this part manually, by calling setSelectedParts directly.
Sets the selected state of the respective axis parts described by SelectablePart. When a part is selected, it uses a different pen/font.
The entire selection mechanism for axes is handled automatically when QCustomPlot::setInteractions contains iSelectAxes. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state of a part, independent of the setSelectableParts setting.
emits the selectionChanged signal when selected is different from the previous selection state.
Sets the style for the lower axis ending. See the documentation of QCPLineEnding for available styles.
For horizontal axes, this method refers to the left ending, for vertical axes the bottom ending. Note that this meaning does not change when the axis range is reversed with setRangeReversed.
Sets the style for the upper axis ending. See the documentation of QCPLineEnding for available styles.
For horizontal axes, this method refers to the right ending, for vertical axes the top ending. Note that this meaning does not change when the axis range is reversed with setRangeReversed.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
Returns the orientation of the axis. The axis orientation (horizontal or vertical) is deduced from the axis type (left, top, right or bottom).
void QCPAxis::moveRange
(
double
diff
)
If the scale type (setScaleType) is stLinear, diff is added to the lower and upper bounds of the range. The range is simply moved by diff.
If the scale type is stLogarithmic, the range bounds are multiplied by diff. This corresponds to an apparent "linear" move in logarithmic scaling by a distance of log(diff).
void QCPAxis::scaleRange
(
double
factor,
double
center
)
Scales the range of this axis by factor around the coordinate center. For example, if factor is 2.0, center is 1.0, then the axis range will double its size, and the point at coordinate 1.0 won't have changed its position in the QCustomPlot widget (i.e. coordinates around 1.0 will have moved symmetrically closer to 1.0).
Scales the range of this axis to have a certain scale ratio to otherAxis. The scaling will be done around the center of the current axis range.
For example, if ratio is 1, this axis is the yAxis and otherAxis is xAxis, graphs plotted with those axes will appear in a 1:1 aspect ratio, independent of the aspect ratio the axis rect has.
This is an operation that changes the range of this axis once, it doesn't fix the scale ratio indefinitely. Note that calling this function in the constructor of the QCustomPlot's parent won't have the desired effect, since the widget dimensions aren't defined yet, and a resizeEvent will follow.
void QCPAxis::rescale
(
bool
onlyVisiblePlottables = false
)
Changes the axis range such that all plottables associated with this axis are fully visible in that dimension.
Returns the part of the axis that is hit by pos (in pixels). The return value of this function is independent of the user-selectable parts defined with setSelectableParts. Further, this function does not change the current selection state of the axis.
If the axis is not visible (setVisible), this function always returns spNone.
Returns a list of all the items that are associated with this axis. An item is considered associated with an axis if at least one of its positions uses the axis as key or value axis.
If you only want static ticks you probably don't need this signal, since you can just set the tick vector (and possibly tick label vector) once. However, if you want to provide ticks (and maybe labels) dynamically, e.g. depending on the current axis range, connect a slot to this signal and set the vector/vectors there.
This signal is emitted when the range of this axis has changed. You can connect it to the setRange slot of another axis to communicate the new range to the other axis, in order for it to be synchronized.
Additionally to the new range, this signal also provides the previous range held by the axis as oldRange.
void QCPAxis::selectionChanged
(
const QCPAxis::SelectableParts &
parts
)
signal
This signal is emitted when the selection state of this axis has changed, either by user interaction or by a direct call to setSelectedParts.
void QCPAxis::setupTickVectors
(
)
protectedvirtual
This function is called to prepare the tick vector, sub tick vector and tick label vector. If setAutoTicks is set to true, appropriate tick values are determined automatically via generateAutoTicks. If it's set to false, the signal ticksRequest is emitted, which can be used to provide external tick positions. Then the sub tick vectors and tick label vectors are created.
void QCPAxis::generateAutoTicks
(
)
protectedvirtual
If setAutoTicks is set to true, this function is called by setupTickVectors to generate reasonable tick positions (and subtick count). The algorithm tries to create approximately mAutoTickCount ticks (set via setAutoTickCount).
If the scale is logarithmic, setAutoTickCount is ignored, and one tick is generated at every power of the current logarithm base, set via setScaleLogBase.
int QCPAxis::calculateAutoSubTickCount
(
double
tickStep
)
const
protectedvirtual
Called by generateAutoTicks when setAutoSubTicks is set to true. Depending on the tickStep between two major ticks on the axis, a different number of sub ticks is appropriate. For Example taking 4 sub ticks for a tickStep of 1 makes more sense than taking 5 sub ticks, because this corresponds to a sub tick step of 0.2, instead of the less intuitive 0.16667. Note that a subtick count of 4 means dividing the major tick step into 5 sections.
This is implemented by a hand made lookup for integer tick steps as well as fractional tick steps with a fractional part of (approximately) 0.5. If a tick step is different (i.e. has no fractional part close to 0.5), the currently set sub tick count (setSubTickCount) is returned.
int QCPAxis::calculateMargin
(
)
protectedvirtual
Returns the appropriate outward margin for this axis. It is needed if QCPAxisRect::setAutoMargins is set to true on the parent axis rect. An axis with axis type atLeft will return an appropriate left margin, atBottom will return an appropriate bottom margin and so forth. For the calculation, this function goes through similar steps as draw, so changing one function likely requires the modification of the other one as well.
The margin consists of the outward tick length, tick label padding, tick label size, label padding, label size, and padding.
The margin is cached internally, so repeated calls while leaving the axis range, fonts, etc. unchanged are very fast.
Draws a single tick label with the provided painter, utilizing the internal label cache to significantly speed up drawing of labels that were drawn in previous calls. The tick label is always bound to an axis, the distance to the axis is controllable via distanceToAxis in pixels. The pixel position in the axis direction is passed in the position parameter. Hence for the bottom axis, position would indicate the horizontal pixel position (not coordinate), at which the label should be drawn.
In order to later draw the axis label in a place that doesn't overlap with the tick labels, the largest tick label size is needed. This is acquired by passing a tickLabelsSize to the drawTickLabel calls during the process of drawing all tick labels of one axis. In every call, tickLabelsSize is expanded, if the drawn label exceeds the value tickLabelsSize currently holds.
The label is drawn with the font and pen that are currently set on the painter. To draw superscripted powers, the font is temporarily made smaller by a fixed factor (see getTickLabelData).
Draws the tick label specified in labelData with painter at the pixel positions x and y. This function is used by placeTickLabel to create new tick labels for the cache, or to directly draw the labels on the QCustomPlot surface when label caching is disabled, i.e. when QCP::phCacheLabels plotting hint is not set.
Transforms the passed text and font to a tickLabelData structure that can then be further processed by getTickLabelDrawOffset and drawTickLabel. It splits the text into base and exponent if necessary (see setNumberFormat) and calculates appropriate bounding boxes.
Calculates the offset at which the top left corner of the specified tick label shall be drawn. The offset is relative to a point right next to the tick the label belongs to.
This function is thus responsible for e.g. centering tick labels under ticks and positioning them appropriately when they are rotated.
void QCPAxis::getMaxTickLabelSize
(
const QFont &
font,
const QString &
text,
QSize *
tickLabelsSize
)
const
protectedvirtual
Simulates the steps done by placeTickLabel by calculating bounding boxes of the text label to be drawn, depending on number format etc. Since only the largest tick label is wanted for the margin calculation, the passed tickLabelsSize is only expanded, if it's currently set to a smaller width/height.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Returns via lowIndex and highIndex, which ticks in the current tick vector are visible in the current range. The return values are indices of the tick vector, not the positions of the ticks themselves.
The actual use of this function is when an external tick vector is provided, since it might exceed far beyond the currently displayed range, and would cause unnecessary calculations e.g. of subticks.
If all ticks are outside the axis range, an inverted range is returned, i.e. highIndex will be smaller than lowIndex. There is one case, where this function returns indices that are not really visible in the current axis range: When the tick spacing is larger than the axis range size and one tick is below the axis range and the next tick is already above the axis range. Because in such cases it is usually desirable to know the tick pair, to draw proper subticks.
double QCPAxis::baseLog
(
double
value
)
const
protected
A log function with the base mScaleLogBase, used mostly for coordinate transforms in logarithmic scales with arbitrary log base. Uses the buffered mScaleLogBaseLogInv for faster calculation. This is set to 1.0/qLn(mScaleLogBase) in setScaleLogBase.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
This is an abstract base class for layout elements whose main purpose is to define the position and size of other child layout elements. In most cases, layouts don't draw anything themselves (but there are exceptions to this, e.g. QCPLegend).
QCPLayout introduces a common interface for accessing and manipulating the child elements. Those functions are most notably elementCount, elementAt, takeAt, take, simplify, removeAt, remove and clear. Individual subclasses may add more functions to this interface which are more specialized to the form of the layout. For example, QCPLayoutGrid adds functions that take row and column indices to access cells of the layout grid more conveniently.
Since this is an abstract base class, you can't instantiate it directly. Rather use one of its subclasses like QCPLayoutGrid or QCPLayoutInset.
For a general introduction to the layout system, see the dedicated documentation page The Layout System.
Constructor & Destructor Documentation
QCPLayout::QCPLayout
(
)
explicit
Creates an instance of QCPLayoutElement and sets default values. Note that since QCPLayoutElement is an abstract base class, it can't be instantiated directly.
Member Function Documentation
void QCPLayout::update
(
)
virtual
First calls the QCPLayoutElement::update base class implementation to update the margins on this layout.
Then calls updateLayout which subclasses reimplement to reposition and resize their cells.
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
Returns the element in the cell with the given index. If index is invalid, returns 0.
Note that even if index is valid, the respective cell may be empty in some layouts (e.g. QCPLayoutGrid), so this function may return 0 in those cases. You may use this function to check whether a cell is empty or not.
Removes the element with the given index from the layout and returns it.
If the index is invalid or the cell with that index is empty, returns 0.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Removes the specified element from the layout and returns true on success.
If the element isn't in this layout, returns false.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Subclasses reimplement this method to update the position and sizes of the child elements/cells via calling their QCPLayoutElement::setOuterRect. The default implementation does nothing.
The geometry used as a reference is the inner rect of this layout. Child elements should stay within that rect.
getSectionSizes may help with the reimplementation of this function.
Subclasses call this method to report changed (minimum/maximum) size constraints.
If the parent of this layout is again a QCPLayout, forwards the call to the parent's sizeConstraintsChanged. If the parent is a QWidget (i.e. is the QCustomPlot::plotLayout of QCustomPlot), calls QWidget::updateGeometry, so if the QCustomPlot widget is inside a Qt QLayout, it may update itself and resize cells accordingly.
This method is used by subclass specific methods that add elements to the layout. Note that this method only changes properties in el. The removal from the old layout and the insertion into the new layout must be done additionally.
This method is used by subclass specific methods that remove elements from the layout (e.g. take or takeAt). Note that this method only changes properties in el. The removal from the old layout must be done additionally.
QVector< int > QCPLayout::getSectionSizes
(
QVector< int >
maxSizes,
QVector< int >
minSizes,
QVector< double >
stretchFactors,
int
totalSize
)
const
protected
This is a helper function for the implementation of updateLayout in subclasses.
It calculates the sizes of one-dimensional sections with provided constraints on maximum section sizes, minimum section sizes, relative stretch factors and the final total size of all sections.
The QVector entries refer to the sections. Thus all QVectors must have the same size.
maxSizes gives the maximum allowed size of each section. If there shall be no maximum size imposed, set all vector values to Qt's QWIDGETSIZE_MAX.
minSizes gives the minimum allowed size of each section. If there shall be no minimum size imposed, set all vector values to zero. If the minSizes entries add up to a value greater than totalSize, sections will be scaled smaller than the proposed minimum sizes. (In other words, not exceeding the allowed total size is taken to be more important than not going below minimum section sizes.)
stretchFactors give the relative proportions of the sections to each other. If all sections shall be scaled equally, set all values equal. If the first section shall be double the size of each individual other section, set the first number of stretchFactors to double the value of the other individual values (e.g. {2, 1, 1, 1}).
totalSize is the value that the final section sizes will add up to. Due to rounding, the actual sum may differ slightly. If you want the section sizes to sum up to exactly that value, you could distribute the remaining difference on the sections.
The return value is a QVector containing the section sizes.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Sets the margin group of the specified margin sides.
Margin groups allow synchronizing specified margins across layout elements, see the documentation of QCPMarginGroup.
To unset the margin group of sides, set group to 0.
Note that margin groups only work for margin sides that are set to automatic (setAutoMargins).
QSize QCPLayoutElement::minimumSizeHint
(
)
const
virtualinherited
Returns the minimum size this layout element (the inner rect) may be compressed to.
if a minimum size (setMinimumSize) was not set manually, parent layouts consult this function to determine the minimum allowed size of this layout element. (A manual minimum size is considered set if it is non-zero.)
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Layout elements are sensitive to events inside their outer rect. If pos is within the outer rect, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. However, layout elements are not selectable by default. So if onlySelectable is true, -1.0 is returned.
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
This function applies the default antialiasing setting to the specified painter, using the function applyAntialiasingHint. It is the antialiasing state the painter is put in, when draw is called on the layerable. If the layerable has multiple entities whose antialiasing setting may be specified individually, this function should set the antialiasing state of the most prominent entity. In this case however, the draw function usually calls the specialized versions of this function before drawing each entity, effectively overriding the setting of the default antialiasing hint.
Second example:QCPItemLine consists only of a line so there is only one antialiasing setting which can be controlled with QCPItemLine::setAntialiased. (This function is inherited by all layerables. The specialized functions, as seen on QCPGraph, must be added explicitly to the respective layerable subclass.) Consequently it only has the normal QCPItemLine::applyDefaultAntialiasingHint. The QCPItemLine::draw function doesn't need to care about setting any antialiasing states, because the default antialiasing hint is already set on the painter when the draw function is called, and that's the state it wants to draw the line with.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Constructs a QCustomPlot and sets reasonable default values.
Member Function Documentation
QRect QCustomPlot::viewport
(
)
const
inline
Returns the viewport rect of this QCustomPlot instance. The viewport is the area the plot is drawn in, all mechanisms, e.g. margin caluclation take the viewport to be the outer border of the plot. The viewport normally is the rect() of the QCustomPlot widget, i.e. a rect with top left (0, 0) and size of the QCustomPlot widget.
Don't confuse the viewport with the axis rect (QCustomPlot::axisRect). An axis rect is typically an area enclosed by four axes, where the graphs/plottables are drawn in. The viewport is larger and contains also the axes themselves, their tick numbers, their labels, the plot title etc.
Only when saving to a file (see savePng, savePdf etc.) the viewport is temporarily modified to allow saving plots with sizes independent of the current widget size.
Returns the top level layout of this QCustomPlot instance. It is a QCPLayoutGrid, initially containing just one cell with the main QCPAxisRect inside.
void QCustomPlot::setViewport
(
const QRect &
rect
)
Sets the viewport of this QCustomPlot. The Viewport is the area that the top level layout (QCustomPlot::plotLayout()) uses as its rect. Normally, the viewport is the entire widget rect.
This function is used to allow arbitrary size exports with toPixmap, savePng, savePdf, etc. by temporarily changing the viewport size.
void QCustomPlot::setBackground
(
const QPixmap &
pm
)
Sets pm as the viewport background pixmap (see setViewport). The pixmap is always drawn below all other objects in the plot.
For cases where the provided pixmap doesn't have the same size as the viewport, scaling can be enabled with setBackgroundScaled and the scaling mode (whether and how the aspect ratio is preserved) can be set with setBackgroundScaledMode. To set all these options in one call, consider using the overloaded version of this function.
If a background brush was set with setBackground(const QBrush &brush), the viewport will first be filled with that brush, before drawing the background pixmap. This can be useful for background pixmaps with translucent areas.
Sets the background brush of the viewport (see setViewport).
Before drawing everything else, the background is filled with brush. If a background pixmap was set with setBackground(const QPixmap &pm), this brush will be used to fill the viewport before the background pixmap is drawn. This can be useful for background pixmaps with translucent areas.
Set brush to Qt::NoBrush or Qt::Transparent to leave background transparent. This can be useful for exporting to image formats which support transparency, e.g. savePng.
Sets whether the viewport background pixmap shall be scaled to fit the viewport. If scaled is set to true, control whether and how the aspect ratio of the original pixmap is preserved with setBackgroundScaledMode.
Note that the scaled version of the original pixmap is buffered, so there is no performance penalty on replots. (Except when the viewport dimensions are changed continuously.)
If scaling of the viewport background pixmap is enabled (setBackgroundScaled), use this function to define whether and how the aspect ratio of the original pixmap is preserved.
Sets which elements are forcibly drawn antialiased as an or combination of QCP::AntialiasedElement.
This overrides the antialiasing settings for whole element groups, normally controlled with the setAntialiasing function on the individual elements. If an element is neither specified in setAntialiasedElements nor in setNotAntialiasedElements, the antialiasing setting on each individual element instance is used.
Sets which elements are forcibly drawn not antialiased as an or combination of QCP::AntialiasedElement.
This overrides the antialiasing settings for whole element groups, normally controlled with the setAntialiasing function on the individual elements. If an element is neither specified in setAntialiasedElements nor in setNotAntialiasedElements, the antialiasing setting on each individual element instance is used.
Plottable selection is controlled by QCP::iSelectPlottables. If QCP::iSelectPlottables is set, the user may select plottables (graphs, curves, bars,...) by clicking on them or in their vicinity (setSelectionTolerance). Whether the user can actually select a plottable can further be restricted with the QCPAbstractPlottable::setSelectable function on the specific plottable. To find out whether a specific plottable is selected, call QCPAbstractPlottable::selected(). To retrieve a list of all currently selected plottables, call selectedPlottables. If you're only interested in QCPGraphs, you may use the convenience function selectedGraphs.
Item selection is controlled by QCP::iSelectItems. If QCP::iSelectItems is set, the user may select items (QCPItemLine, QCPItemText,...) by clicking on them or in their vicinity. To find out whether a specific item is selected, call QCPAbstractItem::selected(). To retrieve a list of all currently selected items, call selectedItems.
Axis selection is controlled with QCP::iSelectAxes. If QCP::iSelectAxes is set, the user may select parts of the axes by clicking on them. What parts exactly (e.g. Axis base line, tick labels, axis label) are selectable can be controlled via QCPAxis::setSelectableParts for each axis. To retrieve a list of all axes that currently contain selected parts, call selectedAxes. Which parts of an axis are selected, can be retrieved with QCPAxis::selectedParts().
Legend selection is controlled with QCP::iSelectLegend. If this is set, the user may select the legend itself or individual items by clicking on them. What parts exactly are selectable can be controlled via QCPLegend::setSelectableParts. To find out whether the legend or any of its child items are selected, check the value of QCPLegend::selectedParts. To find out which child items are selected, call QCPLegend::selectedItems.
All other selectable elements The selection of all other selectable objects (e.g. QCPPlotTitle, or your own layerable subclasses) is controlled with QCP::iSelectOther. If set, the user may select those objects by clicking on them. To find out which are currently selected, you need to check their selected state explicitly.
If the selection state has changed by user interaction, the selectionChangedByUser signal is emitted. Each selectable object additionally emits an individual selectionChanged signal whenever their selection state has changed, i.e. not only by user interaction.
In addition to the selection mechanism presented here, QCustomPlot always emits corresponding signals, when an object is clicked or double clicked. see plottableClick and plottableDoubleClick for example.
Sets the tolerance that is used to decide whether a click selects an object (e.g. a plottable) or not.
If the user clicks in the vicinity of the line of e.g. a QCPGraph, it's only regarded as a potential selection when the minimum distance between the click position and the graph line is smaller than pixels. Objects that are defined by an area (e.g. QCPBars) only react to clicks directly inside the area and ignore this selection tolerance. In other words, it only has meaning for parts of objects that are too thin to exactly hit with a click and thus need such a tolerance.
Sets whether antialiasing is disabled for this QCustomPlot while the user is dragging axes ranges. If many objects, especially plottables, are drawn antialiased, this greatly improves performance during dragging. Thus it creates a more responsive user experience. As soon as the user stops dragging, the last replot is done with normal antialiasing, to restore high image quality.
Sets the keyboard modifier that will be recognized as multi-select-modifier.
If QCP::iMultiSelect is specified in setInteractions, the user may select multiple objects by clicking on them one after the other while holding down modifier.
By default the multi-select-modifier is set to Qt::ControlModifier.
Adds the specified plottable to the plot and, if setAutoAddPlottableToLegend is enabled, to the legend (QCustomPlot::legend). QCustomPlot takes ownership of the plottable.
Returns true on success, i.e. when plottable isn't already in the plot and the parent plot of plottable is this QCustomPlot (the latter is controlled by what axes were passed in the plottable's constructor).
Returns the plottable at the pixel position pos. Plottables that only consist of single lines (like graphs) have a tolerance band around them, see setSelectionTolerance. If multiple plottables come into consideration, the one closest to pos is returned.
Creates a new graph inside the plot. If keyAxis and valueAxis are left unspecified (0), the bottom (xAxis) is used as key and the left (yAxis) is used as value axis. If specified, keyAxis and valueAxis must reside in this QCustomPlot.
keyAxis will be used as key axis (typically "x") and valueAxis as value axis (typically "y") for the graph.
Returns a pointer to the newly created graph, or 0 if adding the graph failed.
Removes the specified graph from the plot and, if necessary, from the QCustomPlot::legend. If any other graphs in the plot have a channel fill set towards the removed graph, the channel fill property of those graphs is reset to zero (no channel fill).
Returns the item at the pixel position pos. Items that only consist of single lines (e.g. QCPItemLine or QCPItemCurve) have a tolerance band around them, see setSelectionTolerance. If multiple items come into consideration, the one closest to pos is returned.
Returns the layer that is set as current layer (see setCurrentLayer).
bool QCustomPlot::setCurrentLayer
(
const QString &
name
)
Sets the layer with the specified name to be the current layer. All layerables (QCPLayerable), e.g. plottables and items, are created on the current layer.
Returns true on success, i.e. if there is a layer with the specified name in the QCustomPlot.
Adds a new layer to this QCustomPlot instance. The new layer will have the name name, which must be unique. Depending on insertMode, it is positioned either below or above otherLayer.
Returns true on success, i.e. if there is no other layer named name and otherLayer is a valid layer inside this QCustomPlot.
If otherLayer is 0, the highest layer in the QCustomPlot will be used.
For an explanation of what layers are in QCustomPlot, see the documentation of QCPLayer.
Removes the specified layer and returns true on success.
All layerables (e.g. plottables and items) on the removed layer will be moved to the layer below layer. If layer is the bottom layer, the layerables are moved to the layer above. In both cases, the total rendering order of all layerables in the QCustomPlot is preserved.
If layer is the current layer (setCurrentLayer), the layer below (or above, if bottom layer) becomes the new current layer.
It is not possible to remove the last layer of the plot.
Initially, only one axis rect (with index 0) exists in the plot. If multiple axis rects were added, all of them may be accessed with this function in a linear fashion (even when they are nested in a layout hierarchy or inside other axis rects via QCPAxisRect::insetLayout).
Returns the layout element at pixel position pos. If there is no element at that position, returns 0.
Only visible elements are used. If QCPLayoutElement::setVisible on the element itself or on any of its parent elements is set to false, it will not be considered.
Rescales the axes such that all plottables (like graphs) in the plot are fully visible.
if onlyVisiblePlottables is set to true, only the plottables that have their visibility set to true (QCPLayerable::setVisible), will be used to rescale the axes.
Deselects all layerables (plottables, items, axes, legends,...) of the QCustomPlot.
Since calling this function is not a user interaction, this does not emit the selectionChangedByUser signal. The individual selectionChanged signals are emitted though, if the objects were previously selected.
Saves a PDF with the vectorized plot to the file fileName. The axis ratio as well as the scale of texts and lines will be derived from the specified width and height. This means, the output will look like the normal on-screen output of a QCustomPlot widget with the corresponding pixel width and height. If either width or height is zero, the exported image will have the same dimensions as the QCustomPlot widget currently has.
noCosmeticPen disables the use of cosmetic pens when drawing to the PDF file. Cosmetic pens are pens with numerical width 0, which are always drawn as a one pixel wide line, no matter what zoom factor is set in the PDF-Viewer. For more information about cosmetic pens, see the QPainter and QPen documentation.
The objects of the plot will appear in the current selection state. If you don't want any selected objects to be painted in their selected look, deselect everything with deselectAll before calling this function.
Returns true on success.
Warning
If you plan on editing the exported PDF file with a vector graphics editor like Inkscape, it is advised to set noCosmeticPen to true to avoid losing those cosmetic lines (which might be quite many, because cosmetic pens are the default for e.g. axes and tick marks).
If calling this function inside the constructor of the parent of the QCustomPlot widget (i.e. the MainWindow constructor, if QCustomPlot is inside the MainWindow), always provide explicit non-zero widths and heights. If you leave width or height as 0 (default), this function uses the current width and height of the QCustomPlot widget. However, in Qt, these aren't defined yet inside the constructor, so you would get an image that has strange widths/heights.
Note
On Android systems, this method does nothing and issues an according qDebug warning message.
Saves a PNG image file to fileName on disc. The output plot will have the dimensions width and height in pixels. If either width or height is zero, the exported image will have the same dimensions as the QCustomPlot widget currently has. Line widths and texts etc. are not scaled up when larger widths/heights are used. If you want that effect, use the scale parameter.
For example, if you set both width and height to 100 and scale to 2, you will end up with an image file of size 200*200 in which all graphical elements are scaled up by factor 2 (line widths, texts, etc.). This scaling is not done by stretching a 100*100 image, the result will have full 200*200 pixel resolution.
If you use a high scaling factor, it is recommended to enable antialiasing for all elements via temporarily setting QCustomPlot::setAntialiasedElements to QCP::aeAll as this allows QCustomPlot to place objects with sub-pixel accuracy.
Warning
If calling this function inside the constructor of the parent of the QCustomPlot widget (i.e. the MainWindow constructor, if QCustomPlot is inside the MainWindow), always provide explicit non-zero widths and heights. If you leave width or height as 0 (default), this function uses the current width and height of the QCustomPlot widget. However, in Qt, these aren't defined yet inside the constructor, so you would get an image that has strange widths/heights.
The objects of the plot will appear in the current selection state. If you don't want any selected objects to be painted in their selected look, deselect everything with deselectAll before calling this function.
If you want the PNG to have a transparent background, call setBackground(const QBrush &brush) with no brush (Qt::NoBrush) or a transparent color (Qt::transparent), before saving.
PNG compression can be controlled with the quality parameter which must be between 0 and 100 or -1 to use the default setting.
Returns true on success. If this function fails, most likely the PNG format isn't supported by the system, see Qt docs about QImageWriter::supportedImageFormats().
Saves a JPG image file to fileName on disc. The output plot will have the dimensions width and height in pixels. If either width or height is zero, the exported image will have the same dimensions as the QCustomPlot widget currently has. Line widths and texts etc. are not scaled up when larger widths/heights are used. If you want that effect, use the scale parameter.
For example, if you set both width and height to 100 and scale to 2, you will end up with an image file of size 200*200 in which all graphical elements are scaled up by factor 2 (line widths, texts, etc.). This scaling is not done by stretching a 100*100 image, the result will have full 200*200 pixel resolution.
If you use a high scaling factor, it is recommended to enable antialiasing for all elements via temporarily setting QCustomPlot::setAntialiasedElements to QCP::aeAll as this allows QCustomPlot to place objects with sub-pixel accuracy.
Warning
If calling this function inside the constructor of the parent of the QCustomPlot widget (i.e. the MainWindow constructor, if QCustomPlot is inside the MainWindow), always provide explicit non-zero widths and heights. If you leave width or height as 0 (default), this function uses the current width and height of the QCustomPlot widget. However, in Qt, these aren't defined yet inside the constructor, so you would get an image that has strange widths/heights.
The objects of the plot will appear in the current selection state. If you don't want any selected objects to be painted in their selected look, deselect everything with deselectAll before calling this function.
JPG compression can be controlled with the quality parameter which must be between 0 and 100 or -1 to use the default setting.
Returns true on success. If this function fails, most likely the JPG format isn't supported by the system, see Qt docs about QImageWriter::supportedImageFormats().
Saves a BMP image file to fileName on disc. The output plot will have the dimensions width and height in pixels. If either width or height is zero, the exported image will have the same dimensions as the QCustomPlot widget currently has. Line widths and texts etc. are not scaled up when larger widths/heights are used. If you want that effect, use the scale parameter.
For example, if you set both width and height to 100 and scale to 2, you will end up with an image file of size 200*200 in which all graphical elements are scaled up by factor 2 (line widths, texts, etc.). This scaling is not done by stretching a 100*100 image, the result will have full 200*200 pixel resolution.
If you use a high scaling factor, it is recommended to enable antialiasing for all elements via temporarily setting QCustomPlot::setAntialiasedElements to QCP::aeAll as this allows QCustomPlot to place objects with sub-pixel accuracy.
Warning
If calling this function inside the constructor of the parent of the QCustomPlot widget (i.e. the MainWindow constructor, if QCustomPlot is inside the MainWindow), always provide explicit non-zero widths and heights. If you leave width or height as 0 (default), this function uses the current width and height of the QCustomPlot widget. However, in Qt, these aren't defined yet inside the constructor, so you would get an image that has strange widths/heights.
The objects of the plot will appear in the current selection state. If you don't want any selected objects to be painted in their selected look, deselect everything with deselectAll before calling this function.
Returns true on success. If this function fails, most likely the BMP format isn't supported by the system, see Qt docs about QImageWriter::supportedImageFormats().
Saves the plot to a rastered image file fileName in the image format format. The plot is sized to width and height in pixels and scaled with scale. (width 100 and scale 2.0 lead to a full resolution file with width 200.) If the format supports compression, quality may be between 0 and 100 to control it.
Returns true on success. If this function fails, most likely the given format isn't supported by the system, see Qt docs about QImageWriter::supportedImageFormats().
The plot is sized to width and height in pixels. If the painter's scale is not 1.0, the resulting plot will appear scaled accordingly.
Note
If you are restricted to using a QPainter (instead of QCPPainter), create a temporary QPicture and open a QCPPainter on it. Then call toPainter with this QCPPainter. After ending the paint operation on the picture, draw it with the QPainter. This will reproduce the painter actions the QCPPainter took, with a QPainter.
Causes a complete replot into the internal buffer. Finally, update() is called, to redraw the buffer on the QCustomPlot widget surface. This is the method that must be called to make changes, for example on the axis ranges or data points of graphs, visible.
Under a few circumstances, QCustomPlot causes a replot by itself. Those are resize events of the QCustomPlot widget and user interactions (object selection and range dragging/zooming).
Before the replot happens, the signal beforeReplot is emitted. After the replot, afterReplot is emitted. It is safe to mutually connect the replot slot with any of those two signals on two QCustomPlots to make them replot synchronously, it won't cause an infinite recursion.
void QCustomPlot::mouseDoubleClick
(
QMouseEvent *
event
)
signal
This signal is emitted when the QCustomPlot receives a mouse double click event.
void QCustomPlot::mousePress
(
QMouseEvent *
event
)
signal
This signal is emitted when the QCustomPlot receives a mouse press event.
It is discouraged to change the drag-axes with QCPAxisRect::setRangeDragAxes here, because the dragging starting point was saved the moment the mouse was pressed. Thus it only has a meaning for the range drag axes that were set at that moment. If you want to change the drag axes, consider doing this in the mousePress signal instead.
void QCustomPlot::mouseRelease
(
QMouseEvent *
event
)
signal
This signal is emitted when the QCustomPlot receives a mouse release event.
This signal is emitted when a legend (item) is clicked.
event is the mouse event that caused the click, legend is the legend that received the click and item is the legend item that received the click. If only the legend and no item is clicked, item is 0. This happens for a click inside the legend padding or the space between two items.
This signal is emitted when a legend (item) is double clicked.
event is the mouse event that caused the click, legend is the legend that received the click and item is the legend item that received the click. If only the legend and no item is clicked, item is 0. This happens for a click inside the legend padding or the space between two items.
This signal is emitted after the user has changed the selection in the QCustomPlot, e.g. by clicking. It is not emitted when the selection state of an object has changed programmatically by a direct call to setSelected() on an object or by calling deselectAll.
In addition to this signal, selectable objects also provide individual signals, for example QCPAxis::selectionChanged or QCPAbstractPlottable::selectionChanged. Note that those signals are emitted even if the selection state is changed programmatically.
See the documentation of setInteractions for details about the selection mechanism.
This signal is emitted immediately before a replot takes place (caused by a call to the slot replot).
It is safe to mutually connect the replot slot with this signal on two QCustomPlots to make them replot synchronously, it won't cause an infinite recursion.
This signal is emitted immediately after a replot has taken place (caused by a call to the slot replot).
It is safe to mutually connect the replot slot with this signal on two QCustomPlots to make them replot synchronously, it won't cause an infinite recursion.
Returns a minimum size hint that corresponds to the minimum size of the top level layout (plotLayout). To prevent QCustomPlot from being collapsed to size/width zero, set a minimum size (setMinimumSize) either on the whole QCustomPlot or on any layout elements inside the plot. This is especially important, when placed in a QLayout where other components try to take in as much space as possible (e.g. QMdiArea).
Event handler for when the QCustomPlot widget needs repainting. This does not cause a replot, but draws the internal buffer on the widget surface.
void QCustomPlot::resizeEvent
(
QResizeEvent *
event
)
protectedvirtual
Event handler for a resize of the QCustomPlot widget. Causes the internal buffer to be resized to the new size. The viewport (which becomes the outer rect of mPlotLayout) is resized appropriately. Finally a replot is performed.
void QCustomPlot::mouseDoubleClickEvent
(
QMouseEvent *
event
)
protectedvirtual
Event handler for when a double click occurs. Emits the mouseDoubleClick signal, then emits the specialized signals when certain objecs are clicked (e.g. plottableDoubleClick, axisDoubleClick, etc.). Finally determines the affected layout element and forwards the event to it.
Event handler for when a mouse button is pressed. Emits the mousePress signal. Then determines the affected layout element and forwards the event to it.
Event handler for when the cursor is moved. Emits the mouseMove signal.
If a layout element has mouse capture focus (a mousePressEvent happened on top of the layout element before), the mouseMoveEvent is forwarded to that element.
Event handler for when a mouse button is released. Emits the mouseRelease signal.
If the mouse was moved less than a certain threshold in any direction since the mousePressEvent, it is considered a click which causes the selection mechanism (if activated via setInteractions) to possibly change selection states accordingly. Further, specialized mouse click signals are emitted (e.g. plottableClick, axisClick, etc.)
If a layout element has mouse capture focus (a mousePressEvent happened on top of the layout element before), the mouseReleaseEvent is forwarded to that element.
Event handler for mouse wheel events. First, the mouseWheel signal is emitted. Then determines the affected layout element and forwards the event to it.
This is the main draw function. It draws the entire plot, including background pixmap, with the specified painter. Note that it does not fill the background with the background brush (as the user may specify with setBackground(const QBrush &brush)), this is up to the respective functions calling this method (e.g. replot, toPixmap and toPainter).
This method is used by QCPAxisRect::removeAxis to report removed axes to the QCustomPlot so it may clear its QCustomPlot::xAxis, yAxis, xAxis2 and yAxis2 members accordingly.
This method is used by the QCPLegend destructor to report legend removal to the QCustomPlot so it may clear its QCustomPlot::legend member accordingly.
void QCustomPlot::updateLayerIndices
(
)
const
protected
Assigns all layers their index (QCPLayer::mIndex) in the mLayers list. This method is thus called after every operation that changes the layer indices, like layer removal, layer creation, layer moving.
Returns the layerable at pixel position pos. If onlySelectable is set to true, only those layerables that are selectable will be considered. (Layerable subclasses communicate their selectability via the QCPLayerable::selectTest method, by returning -1.)
selectionDetails is an output parameter that contains selection specifics of the affected layerable. This is useful if the respective layerable shall be given a subsequent QCPLayerable::selectEvent (like in mouseReleaseEvent). selectionDetails usually contains information about which part of the layerable was hit, in multi-part layerables (e.g. QCPAxis::SelectablePart).
If a pixmap was provided via setBackground, this function buffers the scaled version depending on setBackgroundScaled and setBackgroundScaledMode and then draws it inside the viewport with the provided painter. The scaled version is buffered in mScaledBackgroundPixmap to prevent expensive rescaling at every redraw. It is only updated, when the axis rect has changed in a way that requires a rescale of the background pixmap (this is dependant on the setBackgroundScaledMode), or when a differend axis backgroud pixmap was set.
Note that this function does not draw a fill with the background brush (setBackground(const QBrush &brush)) beneath the pixmap.
In QCustomPlot, items are supplemental graphical elements that are neither plottables (QCPAbstractPlottable) nor axes (QCPAxis). While plottables are always tied to two axes and thus plot coordinates, items can also be placed in absolute coordinates independent of any axes. Each specific item has at least one QCPItemPosition member which controls the positioning. Some items are defined by more than one coordinate and thus have two or more QCPItemPosition members (For example, QCPItemRect has topLeft and bottomRight).
This abstract base class defines a very basic interface like visibility and clipping. Since this class is abstract, it can't be instantiated. Use one of the subclasses or create a subclass yourself to create new items.
by default, the positions of the item are bound to the x- and y-Axis of the plot. So we can just set the plot coordinates where the line should start/end:
If we don't want the line to be positioned in plot coordinates but a different coordinate system, e.g. absolute pixel positions on the QCustomPlot surface, we need to change the position type like this:
See the documentation of those functions for what they need to do.
Allowing the item to be positioned
As mentioned, item positions are represented by QCPItemPosition members. Let's assume the new item shall have only one point as its position (as opposed to two like a rect or multiple like a polygon). You then add a public member of type QCPItemPosition like so:
the const makes sure the pointer itself can't be modified from the user of your new item (the QCPItemPosition instance it points to, can be modified, of course). The initialization of this pointer is made easy with the createPosition function. Just assign the return value of this function to each QCPItemPosition in the constructor of your item. createPosition takes a string which is the name of the position, typically this is identical to the variable name. For example, the constructor of QCPItemExample could look like this:
To give your item a visual representation, reimplement the draw function and use the passed QCPPainter to draw the item. You can retrieve the item position in pixel coordinates from the position member(s) via QCPItemPosition::pixelPoint.
To optimize performance you should calculate a bounding rect first (don't forget to take the pen width into account), check whether it intersects the clipRect, and only draw the item at all if this is the case.
The selectTest function
Your implementation of the selectTest function may use the helpers distSqrToLine and rectSelectTest. With these, the implementation of the selection test becomes significantly simpler for most items. See the documentation of selectTest for what the function parameters mean and what the function should return.
Providing anchors
Providing anchors (QCPItemAnchor) starts off like adding a position. First you create a public member, e.g.
and create it in the constructor with the createAnchor function, assigning it a name and an anchor id (an integer enumerating all anchors on the item, you may create an own enum for this). Since anchors can be placed anywhere, relative to the item's position(s), your item needs to provide the position of every anchor with the reimplementation of the anchorPixelPoint(int anchorId) function.
In essence the QCPItemAnchor is merely an intermediary that itself asks your item for the pixel position when anything attached to the anchor needs to know the coordinates.
Base class constructor which initializes base class members.
Member Function Documentation
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Curve example. Blue dotted circles are anchors, solid blue discs are positions.
It has four positions, start and end, which define the end points of the line, and two control points which define the direction the line exits from the start and the direction from which it approaches the end: startDir and endDir.
With setHead and setTail you may set different line ending styles, e.g. to create an arrow.
Often it is desirable for the control points to stay at fixed relative positions to the start/end point. This can be achieved by setting the parent anchor e.g. of startDir simply to start, and then specify the desired pixel offset with QCPItemPosition::setCoords on startDir.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
Returns the pen that should be used for drawing lines. Returns mPen when the item is not selected and mSelectedPen when it is.
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
inherited
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
Returns the pen that should be used for drawing lines. Returns mPen when the item is not selected and mSelectedPen when it is.
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
inherited
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
A layout that places child elements aligned to the border or arbitrarily positioned.
Elements are placed either aligned to the border or at arbitrary position in the area of the layout. Which placement applies is controlled with the InsetPlacement (setInsetPlacement).
Returns the placement type of the element with the specified index.
Qt::Alignment QCPLayoutInset::insetAlignment
(
int
index
)
const
Returns the alignment of the element with the specified index. The alignment only has a meaning, if the inset placement (setInsetPlacement) is ipBorderAligned.
QRectF QCPLayoutInset::insetRect
(
int
index
)
const
Returns the rect of the element with the specified index. The rect only has a meaning, if the inset placement (setInsetPlacement) is ipFree.
If the inset placement (setInsetPlacement) is ipBorderAligned, this function is used to set the alignment of the element with the specified index to alignment.
alignment is an or combination of the following alignment flags: Qt::AlignLeft, Qt::AlignHCenter, Qt::AlighRight, Qt::AlignTop, Qt::AlignVCenter, Qt::AlignBottom. Any other alignment flags will be ignored.
void QCPLayoutInset::setInsetRect
(
int
index,
const QRectF &
rect
)
If the inset placement (setInsetPlacement) is ipFree, this function is used to set the position and size of the element with the specified index to rect.
rect is given in fractions of the whole inset layout rect. So an inset with rect (0, 0, 1, 1) will span the entire layout. An inset with rect (0.6, 0.1, 0.35, 0.35) will be in the top right corner of the layout, with 35% width and height of the parent layout.
Subclasses reimplement this method to update the position and sizes of the child elements/cells via calling their QCPLayoutElement::setOuterRect. The default implementation does nothing.
The geometry used as a reference is the inner rect of this layout. Child elements should stay within that rect.
getSectionSizes may help with the reimplementation of this function.
Returns the element in the cell with the given index. If index is invalid, returns 0.
Note that even if index is valid, the respective cell may be empty in some layouts (e.g. QCPLayoutGrid), so this function may return 0 in those cases. You may use this function to check whether a cell is empty or not.
Removes the element with the given index from the layout and returns it.
If the index is invalid or the cell with that index is empty, returns 0.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Removes the specified element from the layout and returns true on success.
If the element isn't in this layout, returns false.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
The inset layout is sensitive to events only at areas where its child elements are sensitive. If the selectTest method of any of the child elements returns a positive number for pos, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. The inset layout is not selectable itself by default. So if onlySelectable is true, -1.0 is returned.
Adds the specified element to the layout as an inset aligned at the border (setInsetAlignment is initialized with ipBorderAligned). The alignment is set to alignment.
alignment is an or combination of the following alignment flags: Qt::AlignLeft, Qt::AlignHCenter, Qt::AlighRight, Qt::AlignTop, Qt::AlignVCenter, Qt::AlignBottom. Any other alignment flags will be ignored.
Adds the specified element to the layout as an inset with free positioning/sizing (setInsetAlignment is initialized with ipFree). The position and size is set to rect.
rect is given in fractions of the whole inset layout rect. So an inset with rect (0, 0, 1, 1) will span the entire layout. An inset with rect (0.6, 0.1, 0.35, 0.35) will be in the top right corner of the layout, with 35% width and height of the parent layout.
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
Subclasses call this method to report changed (minimum/maximum) size constraints.
If the parent of this layout is again a QCPLayout, forwards the call to the parent's sizeConstraintsChanged. If the parent is a QWidget (i.e. is the QCustomPlot::plotLayout of QCustomPlot), calls QWidget::updateGeometry, so if the QCustomPlot widget is inside a Qt QLayout, it may update itself and resize cells accordingly.
This method is used by subclass specific methods that add elements to the layout. Note that this method only changes properties in el. The removal from the old layout and the insertion into the new layout must be done additionally.
This method is used by subclass specific methods that remove elements from the layout (e.g. take or takeAt). Note that this method only changes properties in el. The removal from the old layout must be done additionally.
QVector< int > QCPLayout::getSectionSizes
(
QVector< int >
maxSizes,
QVector< int >
minSizes,
QVector< double >
stretchFactors,
int
totalSize
)
const
protectedinherited
This is a helper function for the implementation of updateLayout in subclasses.
It calculates the sizes of one-dimensional sections with provided constraints on maximum section sizes, minimum section sizes, relative stretch factors and the final total size of all sections.
The QVector entries refer to the sections. Thus all QVectors must have the same size.
maxSizes gives the maximum allowed size of each section. If there shall be no maximum size imposed, set all vector values to Qt's QWIDGETSIZE_MAX.
minSizes gives the minimum allowed size of each section. If there shall be no minimum size imposed, set all vector values to zero. If the minSizes entries add up to a value greater than totalSize, sections will be scaled smaller than the proposed minimum sizes. (In other words, not exceeding the allowed total size is taken to be more important than not going below minimum section sizes.)
stretchFactors give the relative proportions of the sections to each other. If all sections shall be scaled equally, set all values equal. If the first section shall be double the size of each individual other section, set the first number of stretchFactors to double the value of the other individual values (e.g. {2, 1, 1, 1}).
totalSize is the value that the final section sizes will add up to. Due to rounding, the actual sum may differ slightly. If you want the section sizes to sum up to exactly that value, you could distribute the remaining difference on the sections.
The return value is a QVector containing the section sizes.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Sets the margin group of the specified margin sides.
Margin groups allow synchronizing specified margins across layout elements, see the documentation of QCPMarginGroup.
To unset the margin group of sides, set group to 0.
Note that margin groups only work for margin sides that are set to automatic (setAutoMargins).
QSize QCPLayoutElement::minimumSizeHint
(
)
const
virtualinherited
Returns the minimum size this layout element (the inner rect) may be compressed to.
if a minimum size (setMinimumSize) was not set manually, parent layouts consult this function to determine the minimum allowed size of this layout element. (A manual minimum size is considered set if it is non-zero.)
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
This function applies the default antialiasing setting to the specified painter, using the function applyAntialiasingHint. It is the antialiasing state the painter is put in, when draw is called on the layerable. If the layerable has multiple entities whose antialiasing setting may be specified individually, this function should set the antialiasing state of the most prominent entity. In this case however, the draw function usually calls the specialized versions of this function before drawing each entity, effectively overriding the setting of the default antialiasing hint.
Second example:QCPItemLine consists only of a line so there is only one antialiasing setting which can be controlled with QCPItemLine::setAntialiased. (This function is inherited by all layerables. The specialized functions, as seen on QCPGraph, must be added explicitly to the respective layerable subclass.) Consequently it only has the normal QCPItemLine::applyDefaultAntialiasingHint. The QCPItemLine::draw function doesn't need to care about setting any antialiasing states, because the default antialiasing hint is already set on the painter when the draw function is called, and that's the state it wants to draw the line with.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Represents the visual appearance of scatter points.
This class holds information about shape, color and size of scatter points. In plottables like QCPGraph it is used to store how scatter points shall be drawn. For example, QCPGraph::setScatterStyle takes a QCPScatterStyle instance.
A scatter style consists of a shape (setShape), a line color (setPen) and possibly a fill (setBrush), if the shape provides a fillable area. Further, the size of the shape can be controlled with setSize.
Specifying a scatter style
You can set all these configurations either by calling the respective functions on an instance:
Or you can use one of the various constructors that take different parameter combinations, making it easy to specify a scatter style in a single call, like so:
There are two constructors which leave the pen undefined: QCPScatterStyle() and QCPScatterStyle(ScatterShape shape, double size). If those constructors are used, a call to isPenDefined will return false. It leads to scatter points that inherit the pen from the plottable that uses the scatter style. Thus, if such a scatter style is passed to QCPGraph, the line color of the graph (QCPGraph::setPen) will be used by the scatter points. This makes it very convenient to set up typical scatter settings:
QCPScatterStyle supports drawing custom shapes and arbitrary pixmaps as scatter points.
For custom shapes, you can provide a QPainterPath with the desired shape to the setCustomPath function or call the constructor that takes a painter path. The scatter shape will automatically be set to ssCustom.
For pixmaps, you call setPixmap with the desired QPixmap. Alternatively you can use the constructor that takes a QPixmap. The scatter shape will automatically be set to ssPixmap. Note that setSize does not influence the appearance of the pixmap.
On plottables/items that draw scatters, the sizes of these visualizations (with exception of ssDot and ssPixmap) can be controlled with the setSize function. Scatters are drawn with the pen and brush specified with setPen and setBrush.
Enumerator:
ssNone
no scatter symbols are drawn (e.g. in QCPGraph, data only represented with lines)
ssDot
a single pixel (use ssDisc or ssCircle if you want a round shape with a certain radius)
ssCross
a cross
ssPlus
a plus
ssCircle
a circle
ssDisc
a circle which is filled with the pen's color (not the brush as with ssCircle)
ssSquare
a square
ssDiamond
a diamond
ssStar
a star with eight arms, i.e. a combination of cross and plus
ssTriangle
an equilateral triangle, standing on baseline
ssTriangleInverted
an equilateral triangle, standing on corner
ssCrossSquare
a square with a cross inside
ssPlusSquare
a square with a plus inside
ssCrossCircle
a circle with a cross inside
ssPlusCircle
a circle with a plus inside
ssPeace
a circle, with one vertical and two downward diagonal lines
ssPixmap
a custom pixmap specified by setPixmap, centered on the data point coordinates
ssCustom
custom painter operations are performed per scatter (As QPainterPath, see setCustomPath)
Constructor & Destructor Documentation
QCPScatterStyle::QCPScatterStyle
(
)
Creates a new QCPScatterStyle instance with size set to 6. No shape, pen or brush is defined.
Since the pen is undefined (isPenDefined returns false), the scatter color will be inherited from the plottable that uses this scatter style.
Creates a new QCPScatterStyle instance with shape set to shape, the pen color set to color, and size to size. No brush is defined, i.e. the scatter point will not be filled.
Creates a new QCPScatterStyle instance with shape set to shape, the pen color set to color, the brush color to fill (with a solid pattern), and size to size.
Creates a new QCPScatterStyle instance with shape set to shape, the pen set to pen, the brush to brush, and size to size.
Warning
In some cases it might be tempting to directly use a pen style like Qt::NoPen as pen and a color like Qt::blue as brush. Notice however, that the corresponding call QCPScatterStyle(QCPScatterShape::ssCircle, Qt::NoPen, Qt::blue, 5)
doesn't necessarily lead C++ to use this constructor in some cases, but might mistake Qt::NoPen for a QColor and use the QCPScatterStyle(ScatterShape shape, const QColor &color, const QColor &fill, double size) constructor instead (which will lead to an unexpected look of the scatter points). To prevent this, be more explicit with the parameter types. For example, use QBrush(Qt::blue) instead of just Qt::blue, to clearly point out to the compiler that this constructor is wanted.
QCPScatterStyle::QCPScatterStyle
(
const QPixmap &
pixmap
)
Creates a new QCPScatterStyle instance which will show the specified pixmap. The scatter shape is set to ssPixmap.
QCPScatterStyle::QCPScatterStyle
(
const QPainterPath &
customPath,
const QPen &
pen,
const QBrush &
brush = Qt::NoBrush,
double
size = 6
)
Creates a new QCPScatterStyle instance with a custom shape that is defined via customPath. The scatter shape is set to ssCustom.
The custom shape line will be drawn with pen and filled with brush. The size has a slightly different meaning than for built-in scatter points: The custom path will be drawn scaled by a factor of size/6.0. Since the default size is 6, the custom path will appear at a its natural size by default. To double the size of the path for example, set size to 12.
Member Function Documentation
void QCPScatterStyle::setSize
(
double
size
)
Sets the size (pixel diameter) of the drawn scatter points to size.
Sets the brush that will be used to fill scatter points to brush. Note that not all scatter shapes have fillable areas. For example, ssPlus does not while ssCircle does.
Returns whether a pen has been defined for this scatter style.
The pen is undefined if a constructor is called that does not carry pen as parameter. Those are QCPScatterStyle() and QCPScatterStyle(ScatterShape shape, double size). If the pen is left undefined, the scatter color will be inherited from the plottable that uses this scatter style.
Applies the pen and the brush of this scatter style to painter. If this scatter style has an undefined pen (isPenDefined), sets the pen of painter to defaultPen instead.
This function is used by plottables (or any class that wants to draw scatters) just before a number of scatters with this style shall be drawn with the painter.
By default, QCPAxisRect comes with four axes, at bottom, top, left and right. They can be accessed via axis by providing the respective axis type (QCPAxis::AxisType) and index. If you need all axes in the axis rect, use axes. The top and right axes are set to be invisible initially (QCPAxis::setVisible). To add more axes to a side, use addAxis or addAxes. To remove an axis, use removeAxis.
The axis rect layerable itself only draws a background pixmap or color, if specified (setBackground). It is placed on the "background" layer initially (see QCPLayer for an explanation of the QCustomPlot layer system). The axes that are held by the axis rect can be placed on other layers, independently of the axis rect.
Every axis rect has a child layout of type QCPLayoutInset. It is accessible via insetLayout and can be used to have other layout elements (or even other layouts with multiple elements) hovering inside the axis rect.
Creates a QCPAxisRect instance and sets default values. An axis is added for each of the four sides, the top and right axes are set invisible initially.
Sets pm as the axis background pixmap. The axis background pixmap will be drawn inside the axis rect. Since axis rects place themselves on the "background" layer by default, the axis rect backgrounds are usually drawn below everything else.
For cases where the provided pixmap doesn't have the same size as the axis rect, scaling can be enabled with setBackgroundScaled and the scaling mode (i.e. whether and how the aspect ratio is preserved) can be set with setBackgroundScaledMode. To set all these options in one call, consider using the overloaded version of this function.
Sets brush as the background brush. The axis rect background will be filled with this brush. Since axis rects place themselves on the "background" layer by default, the axis rect backgrounds are usually drawn below everything else.
Sets whether the axis background pixmap shall be scaled to fit the axis rect or not. If scaled is set to true, you may control whether and how the aspect ratio of the original pixmap is preserved with setBackgroundScaledMode.
Note that the scaled version of the original pixmap is buffered, so there is no performance penalty on replots. (Except when the axis rect dimensions are changed continuously.)
If scaling of the axis background pixmap is enabled (setBackgroundScaled), use this function to define whether and how the aspect ratio of the original pixmap passed to setBackground is preserved.
Sets which axis orientation may be range dragged by the user with mouse interaction. What orientation corresponds to which specific axis can be set with setRangeDragAxes(QCPAxis *horizontal, QCPAxis *vertical). By default, the horizontal axis is the bottom axis (xAxis) and the vertical axis is the left axis (yAxis).
To disable range dragging entirely, pass 0 as orientations or remove QCP::iRangeDrag from QCustomPlot::setInteractions. To enable range dragging for both directions, pass Qt::Horizontal | Qt::Vertical as orientations.
Sets which axis orientation may be zoomed by the user with the mouse wheel. What orientation corresponds to which specific axis can be set with setRangeZoomAxes(QCPAxis *horizontal, QCPAxis *vertical). By default, the horizontal axis is the bottom axis (xAxis) and the vertical axis is the left axis (yAxis).
To disable range zooming entirely, pass 0 as orientations or remove QCP::iRangeZoom from QCustomPlot::setInteractions. To enable range zooming for both directions, pass Qt::Horizontal | Qt::Vertical as orientations.
Sets how strong one rotation step of the mouse wheel zooms, when range zoom was activated with setRangeZoom. The two parameters horizontalFactor and verticalFactor provide a way to let the horizontal axis zoom at different rates than the vertical axis. Which axis is horizontal and which is vertical, can be set with setRangeZoomAxes.
When the zoom factor is greater than one, scrolling the mouse wheel backwards (towards the user) will zoom in (make the currently visible range smaller). For zoom factors smaller than one, the same scrolling direction will zoom out.
void QCPAxisRect::setRangeZoomFactor
(
double
factor
)
This is an overloaded function.
Sets both the horizontal and vertical zoom factor.
Adds a new axis with addAxis to each axis rect side specified in types. This may be an or-combination of QCPAxis::AxisType, so axes can be added to multiple sides at once.
Returns the inset layout of this axis rect. It can be used to place other layout elements (or even layouts with multiple other elements) inside/on top of an axis rect.
Convenience function to create an axis on each side that doesn't have any axes yet, and assign the top/right axes the following properties of the bottom/left axes (even if they already existed and weren't created by this function):
Returns a list of all the items that are associated with this axis rect.
An item is considered associated with an axis rect if any of its positions has key or value axis set to an axis that is in this axis rect, or if any of its positions has QCPItemPosition::setAxisRect set to the axis rect, or if the clip axis rect (QCPAbstractItem::setClipAxisRect) is set to this axis rect.
Returns the pixel position of the left border of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
int QCPAxisRect::right
(
)
const
inline
Returns the pixel position of the right border of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
int QCPAxisRect::top
(
)
const
inline
Returns the pixel position of the top border of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
int QCPAxisRect::bottom
(
)
const
inline
Returns the pixel position of the bottom border of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
int QCPAxisRect::width
(
)
const
inline
Returns the pixel width of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
int QCPAxisRect::height
(
)
const
inline
Returns the pixel height of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QSize QCPAxisRect::size
(
)
const
inline
Returns the pixel size of this axis rect. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QPoint QCPAxisRect::topLeft
(
)
const
inline
Returns the top left corner of this axis rect in pixels. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QPoint QCPAxisRect::topRight
(
)
const
inline
Returns the top right corner of this axis rect in pixels. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QPoint QCPAxisRect::bottomLeft
(
)
const
inline
Returns the bottom left corner of this axis rect in pixels. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QPoint QCPAxisRect::bottomRight
(
)
const
inline
Returns the bottom right corner of this axis rect in pixels. Margins are not taken into account here, so the returned value is with respect to the inner rect.
QPoint QCPAxisRect::center
(
)
const
inline
Returns the center of this axis rect in pixels. Margins are not taken into account here, so the returned value is with respect to the inner rect.
void QCPAxisRect::update
(
)
virtual
This method is called automatically upon replot and doesn't need to be called by users of QCPAxisRect.
Calls the base class implementation to update the margins (see QCPLayoutElement::update), and finally passes the rect to the inset layout (insetLayout) and calls its QCPInsetLayout::update function.
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
This function applies the default antialiasing setting to the specified painter, using the function applyAntialiasingHint. It is the antialiasing state the painter is put in, when draw is called on the layerable. If the layerable has multiple entities whose antialiasing setting may be specified individually, this function should set the antialiasing state of the most prominent entity. In this case however, the draw function usually calls the specialized versions of this function before drawing each entity, effectively overriding the setting of the default antialiasing hint.
Second example:QCPItemLine consists only of a line so there is only one antialiasing setting which can be controlled with QCPItemLine::setAntialiased. (This function is inherited by all layerables. The specialized functions, as seen on QCPGraph, must be added explicitly to the respective layerable subclass.) Consequently it only has the normal QCPItemLine::applyDefaultAntialiasingHint. The QCPItemLine::draw function doesn't need to care about setting any antialiasing states, because the default antialiasing hint is already set on the painter when the draw function is called, and that's the state it wants to draw the line with.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
Event handler for when a mouse button is pressed on the axis rect. If the left mouse button is pressed, the range dragging interaction is initialized (the actual range manipulation happens in the mouseMoveEvent).
The mDragging flag is set to true and some anchor points are set that are needed to determine the distance the mouse was dragged in the mouse move/release events later.
Event handler for when the mouse is moved on the axis rect. If range dragging was activated in a preceding mousePressEvent, the range is moved accordingly.
Event handler for mouse wheel events. If rangeZoom is Qt::Horizontal, Qt::Vertical or both, the ranges of the axes defined as rangeZoomHorzAxis and rangeZoomVertAxis are scaled. The center of the scaling operation is the current cursor position inside the axis rect. The scaling factor is dependant on the mouse wheel delta (which direction the wheel was rotated) to provide a natural zooming feel. The Strength of the zoom can be controlled via setRangeZoomFactor.
Note, that event->delta() is usually +/-120 for single rotation steps. However, if the mouse wheel is turned rapidly, many steps may bunch up to one event, so the event->delta() may then be multiples of 120. This is taken into account here, by calculating wheelSteps and using it as exponent of the range zoom factor. This takes care of the wheel direction automatically, by inverting the factor, when the wheel step is negative (f^-1 = 1/f).
Draws the background of this axis rect. It may consist of a background fill (a QBrush) and a pixmap.
If a brush was given via setBackground(const QBrush &brush), this function first draws an according filling inside the axis rect with the provided painter.
Then, if a pixmap was provided via setBackground, this function buffers the scaled version depending on setBackgroundScaled and setBackgroundScaledMode and then draws it inside the axis rect with the provided painter. The scaled version is buffered in mScaledBackgroundPixmap to prevent expensive rescaling at every redraw. It is only updated, when the axis rect has changed in a way that requires a rescale of the background pixmap (this is dependant on the setBackgroundScaledMode), or when a differend axis backgroud pixmap was set.
This function makes sure multiple axes on the side specified with type don't collide, but are distributed according to their respective space requirement (QCPAxis::calculateMargin).
It does this by setting an appropriate offset (QCPAxis::setOffset) on all axes except the one with index zero.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Sets the margin group of the specified margin sides.
Margin groups allow synchronizing specified margins across layout elements, see the documentation of QCPMarginGroup.
To unset the margin group of sides, set group to 0.
Note that margin groups only work for margin sides that are set to automatic (setAutoMargins).
QSize QCPLayoutElement::minimumSizeHint
(
)
const
virtualinherited
Returns the minimum size this layout element (the inner rect) may be compressed to.
if a minimum size (setMinimumSize) was not set manually, parent layouts consult this function to determine the minimum allowed size of this layout element. (A manual minimum size is considered set if it is non-zero.)
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Layout elements are sensitive to events inside their outer rect. If pos is within the outer rect, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. However, layout elements are not selectable by default. So if onlySelectable is true, -1.0 is returned.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Elements can be added to cells via addElement. The grid is expanded if the specified row or column doesn't exist yet. Whether a cell contains a valid layout element can be checked with hasElement, that element can be retrieved with element. If rows and columns that only have empty cells shall be removed, call simplify. Removal of elements is either done by just adding the element to a different layout or by using the QCPLayout interface take or remove.
Subclasses reimplement this method to update the position and sizes of the child elements/cells via calling their QCPLayoutElement::setOuterRect. The default implementation does nothing.
The geometry used as a reference is the inner rect of this layout. Child elements should stay within that rect.
getSectionSizes may help with the reimplementation of this function.
Returns the element in the cell with the given index. If index is invalid, returns 0.
Note that even if index is valid, the respective cell may be empty in some layouts (e.g. QCPLayoutGrid), so this function may return 0 in those cases. You may use this function to check whether a cell is empty or not.
Removes the element with the given index from the layout and returns it.
If the index is invalid or the cell with that index is empty, returns 0.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Removes the specified element from the layout and returns true on success.
If the element isn't in this layout, returns false.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
Returns the minimum size this layout element (the inner rect) may be compressed to.
if a minimum size (setMinimumSize) was not set manually, parent layouts consult this function to determine the minimum allowed size of this layout element. (A manual minimum size is considered set if it is non-zero.)
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Returns the element in the cell in row and column.
Returns 0 if either the row/column is invalid or if the cell is empty. In those cases, a qDebug message is printed. To check whether a cell exists and isn't empty, use hasElement.
Adds the element to cell with row and column. If element is already in a layout, it is first removed from there. If row or column don't exist yet, the layout is expanded accordingly.
Returns true if the element was added successfully, i.e. if the cell at row and column didn't already have an element.
Expands the layout to have newRowCount rows and newColumnCount columns. So the last valid row index will be newRowCount-1, the last valid column index will be newColumnCount-1.
If the current column/row count is already larger or equal to newColumnCount/newRowCount, this function does nothing in that dimension.
Newly created cells are empty, new rows and columns have the stretch factor 1.
Note that upon a call to addElement, the layout is expanded automatically to contain the specified row and column, using this function.
Inserts a new row with empty cells at the row index newIndex. Valid values for newIndex range from 0 (inserts a row at the top) to rowCount (appends a row at the bottom).
Inserts a new column with empty cells at the column index newIndex. Valid values for newIndex range from 0 (inserts a row at the left) to rowCount (appends a row at the right).
Places the minimum column widths and row heights into minColWidths and minRowHeights respectively.
The minimum height of a row is the largest minimum height of any element in that row. The minimum width of a column is the largest minimum width of any element in that column.
Places the maximum column widths and row heights into maxColWidths and maxRowHeights respectively.
The maximum height of a row is the smallest maximum height of any element in that row. The maximum width of a column is the smallest maximum width of any element in that column.
Subclasses call this method to report changed (minimum/maximum) size constraints.
If the parent of this layout is again a QCPLayout, forwards the call to the parent's sizeConstraintsChanged. If the parent is a QWidget (i.e. is the QCustomPlot::plotLayout of QCustomPlot), calls QWidget::updateGeometry, so if the QCustomPlot widget is inside a Qt QLayout, it may update itself and resize cells accordingly.
This method is used by subclass specific methods that add elements to the layout. Note that this method only changes properties in el. The removal from the old layout and the insertion into the new layout must be done additionally.
This method is used by subclass specific methods that remove elements from the layout (e.g. take or takeAt). Note that this method only changes properties in el. The removal from the old layout must be done additionally.
QVector< int > QCPLayout::getSectionSizes
(
QVector< int >
maxSizes,
QVector< int >
minSizes,
QVector< double >
stretchFactors,
int
totalSize
)
const
protectedinherited
This is a helper function for the implementation of updateLayout in subclasses.
It calculates the sizes of one-dimensional sections with provided constraints on maximum section sizes, minimum section sizes, relative stretch factors and the final total size of all sections.
The QVector entries refer to the sections. Thus all QVectors must have the same size.
maxSizes gives the maximum allowed size of each section. If there shall be no maximum size imposed, set all vector values to Qt's QWIDGETSIZE_MAX.
minSizes gives the minimum allowed size of each section. If there shall be no minimum size imposed, set all vector values to zero. If the minSizes entries add up to a value greater than totalSize, sections will be scaled smaller than the proposed minimum sizes. (In other words, not exceeding the allowed total size is taken to be more important than not going below minimum section sizes.)
stretchFactors give the relative proportions of the sections to each other. If all sections shall be scaled equally, set all values equal. If the first section shall be double the size of each individual other section, set the first number of stretchFactors to double the value of the other individual values (e.g. {2, 1, 1, 1}).
totalSize is the value that the final section sizes will add up to. Due to rounding, the actual sum may differ slightly. If you want the section sizes to sum up to exactly that value, you could distribute the remaining difference on the sections.
The return value is a QVector containing the section sizes.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Sets the margin group of the specified margin sides.
Margin groups allow synchronizing specified margins across layout elements, see the documentation of QCPMarginGroup.
To unset the margin group of sides, set group to 0.
Note that margin groups only work for margin sides that are set to automatic (setAutoMargins).
double QCPLayoutElement::selectTest
(
const QPointF &
pos,
bool
onlySelectable,
QVariant *
details = 0
)
const
virtualinherited
Layout elements are sensitive to events inside their outer rect. If pos is within the outer rect, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. However, layout elements are not selectable by default. So if onlySelectable is true, -1.0 is returned.
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
This function applies the default antialiasing setting to the specified painter, using the function applyAntialiasingHint. It is the antialiasing state the painter is put in, when draw is called on the layerable. If the layerable has multiple entities whose antialiasing setting may be specified individually, this function should set the antialiasing state of the most prominent entity. In this case however, the draw function usually calls the specialized versions of this function before drawing each entity, effectively overriding the setting of the default antialiasing hint.
Second example:QCPItemLine consists only of a line so there is only one antialiasing setting which can be controlled with QCPItemLine::setAntialiased. (This function is inherited by all layerables. The specialized functions, as seen on QCPGraph, must be added explicitly to the respective layerable subclass.) Consequently it only has the normal QCPItemLine::applyDefaultAntialiasingHint. The QCPItemLine::draw function doesn't need to care about setting any antialiasing states, because the default antialiasing hint is already set on the painter when the draw function is called, and that's the state it wants to draw the line with.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
A legend item representing a plottable with an icon and the plottable name.
This is the standard legend item for plottables. It displays an icon of the plottable next to the plottable name. The icon is drawn by the respective plottable itself (QCPAbstractPlottable::drawLegendIcon), and tries to give an intuitive symbol for the plottable. For example, the QCPGraph draws a centered horizontal line and/or a single scatter point in the middle.
Legend items of this type are always associated with one plottable (retrievable via the plottable() function and settable with the constructor). You may change the font of the plottable name with setFont. Icon padding and border pen is taken from the parent QCPLegend, see QCPLegend::setIconBorderPen and QCPLegend::setIconTextPadding.
Since QCPLegend is based on QCPLayoutGrid, a legend item itself is just a subclass of QCPLayoutElement. While it could be added to a legend (or any other layout) via the normal layout interface, QCPLegend has specialized functions for handling legend items conveniently, see the documentation of QCPLegend.
Draws the item with painter. The size and position of the drawn legend item is defined by the parent layout (typically a QCPLegend) and the minimumSizeHint and maximumSizeHint of this legend item.
Layout elements are sensitive to events inside their outer rect. If pos is within the outer rect, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. However, layout elements are not selectable by default. So if onlySelectable is true, -1.0 is returned.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This function applies the default antialiasing setting to the specified painter, using the function applyAntialiasingHint. It is the antialiasing state the painter is put in, when draw is called on the layerable. If the layerable has multiple entities whose antialiasing setting may be specified individually, this function should set the antialiasing state of the most prominent entity. In this case however, the draw function usually calls the specialized versions of this function before drawing each entity, effectively overriding the setting of the default antialiasing hint.
Second example:QCPItemLine consists only of a line so there is only one antialiasing setting which can be controlled with QCPItemLine::setAntialiased. (This function is inherited by all layerables. The specialized functions, as seen on QCPGraph, must be added explicitly to the respective layerable subclass.) Consequently it only has the normal QCPItemLine::applyDefaultAntialiasingHint. The QCPItemLine::draw function doesn't need to care about setting any antialiasing states, because the default antialiasing hint is already set on the painter when the draw function is called, and that's the state it wants to draw the line with.
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
The appearance of the box itself is controlled via setPen and setBrush. You may change the width of the box with setWidth in plot coordinates (not pixels).
Analog functions exist for the minimum/maximum-whiskers: setWhiskerPen, setWhiskerBarPen, setWhiskerWidth. The whisker width is the width of the bar at the top (maximum) and bottom (minimum).
The median indicator line has its own pen, setMedianPen.
If the whisker backbone pen is changed, make sure to set the capStyle to Qt::FlatCap. Else, the backbone line might exceed the whisker bars by a few pixels due to the pen cap being not perfectly flat.
The Outlier data points are drawn as normal scatter points. Their look can be controlled with setOutlierStyle
Constructs a statistical box which uses keyAxis as its key axis ("x") and valueAxis as its value axis ("y"). keyAxis and valueAxis must reside in the same QCustomPlot instance and not have the same orientation. If either of these restrictions is violated, a corresponding message is printed to the debug output (qDebug), the construction is not aborted, though.
The constructed statistical box can be added to the plot with QCustomPlot::addPlottable, QCustomPlot then takes ownership of the statistical box.
Member Function Documentation
void QCPStatisticalBox::setKey
(
double
key
)
Sets the key coordinate of the statistical box.
void QCPStatisticalBox::setMinimum
(
double
value
)
Sets the parameter "minimum" of the statistical box plot. This is the position of the lower whisker, typically the minimum measurement of the sample that's not considered an outlier.
Sets the parameter "lower Quartile" of the statistical box plot. This is the lower end of the box. The lower and the upper quartiles are the two statistical quartiles around the median of the sample, they contain 50% of the sample data.
Sets the parameter "median" of the statistical box plot. This is the value of the median mark inside the quartile box. The median separates the sample data in half (50% of the sample data is below/above the median).
Sets the parameter "upper Quartile" of the statistical box plot. This is the upper end of the box. The lower and the upper quartiles are the two statistical quartiles around the median of the sample, they contain 50% of the sample data.
Sets the parameter "maximum" of the statistical box plot. This is the position of the upper whisker, typically the maximum measurement of the sample that's not considered an outlier.
Sets a vector of outlier values that will be drawn as circles. Any data points in the sample that are not within the whiskers (setMinimum, setMaximum) should be considered outliers and displayed as such.
Sets the pen used for drawing the whisker backbone (That's the line parallel to the value axis).
Make sure to set the pen capStyle to Qt::FlatCap to prevent the whisker backbone from reaching a few pixels past the whisker bars, when using a non-zero pen width.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
called by rescaleAxes functions to get the full data key bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
called by rescaleAxes functions to get the full data value bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
Draws the quartile box. box is an output parameter that returns the quartile box (in pixel coordinates) which is used to set the clip rect of the painter before calling drawMedian (so the median doesn't draw outside the quartile box).
The name is the textual representation of this plottable as it is displayed in the legend (QCPLegend). It may contain any UTF-8 characters, including newlines.
void QCPAbstractPlottable::setAntialiasedFill
(
bool
enabled
)
inherited
Sets whether fills of this plottable is drawn antialiased or not.
The brush is used to draw basic fills of the plottable representation in the plot. The Fill can be a color, gradient or texture, see the usage of QBrush.
For example, the QCPGraph subclass draws the fill under the graph with this brush, when it's not set to Qt::NoBrush.
The key axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's value axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
The value axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's key axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
Sets whether the user can (de-)select this plottable by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains iSelectPlottables.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected directly.
Sets whether this plottable is selected or not. When selected, it uses a different pen and brush to draw its lines and fills, see setSelectedPen and setSelectedBrush.
The entire selection mechanism for plottables is handled automatically when QCustomPlot::setInteractions contains iSelectPlottables. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Adds this plottable to the legend of the parent QCustomPlot (QCustomPlot::legend).
Normally, a QCPPlottableLegendItem is created and inserted into the legend. If the plottable needs a more specialized representation in the legend, this function will take this into account and instead create the specialized subclass of QCPAbstractLegendItem.
Returns true on success, i.e. when the legend exists and a legend item associated with this plottable isn't already in the legend.
Rescales the key and value axes associated with this plottable to contain all displayed data, so the whole plottable is visible. If the scaling of an axis is logarithmic, rescaleAxes will make sure not to rescale to an illegal range i.e. a range containing different signs and/or zero. Instead it will stay in the current sign domain and ignore all parts of the plottable that lie outside of that domain.
onlyEnlarge makes sure the ranges are only expanded, never reduced. So it's possible to show multiple plottables in their entirety by multiple calls to rescaleAxes where the first call has onlyEnlarge set to false (the default), and all subsequent set to true.
This signal is emitted when the selection state of this plottable has changed to selected, either by user interaction or by a direct call to setSelected.
QRect QCPAbstractPlottable::clipRect
(
)
const
protectedvirtualinherited
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Convenience function for transforming a key/value pair to pixels on the QCustomPlot surface, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
key and value are transformed to the coodinates in pixels and are written to x and y.
Returns the input as pixel coordinates in a QPointF.
void QCPAbstractPlottable::pixelsToCoords
(
double
x,
double
y,
double &
key,
double &
value
)
const
protectedinherited
Convenience function for transforming a x/y pixel pair on the QCustomPlot surface to plot coordinates, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
x and y are transformed to the plot coodinates and are written to key and value.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
To plot data, assign it with the setData or addData functions.
Graphs are used to display single-valued data. Single-valued means that there should only be one data point per unique key coordinate. In other words, the graph can't have loops. If you do want to plot non-single-valued curves, rather use the QCPCurve plottable.
QCPGraph knows two types of fills: Normal graph fills towards the zero-value-line parallel to the key axis of the graph, and fills between two graphs, called channel fills. To enable a fill, just set a brush with setBrush which is neither Qt::NoBrush nor fully transparent.
By default, a normal fill towards the zero-value-line will be drawn. To set up a channel fill between this graph and another one, call setChannelFillGraph with the other graph as parameter.
Constructs a graph which uses keyAxis as its key axis ("x") and valueAxis as its value axis ("y"). keyAxis and valueAxis must reside in the same QCustomPlot instance and not have the same orientation. If either of these restrictions is violated, a corresponding message is printed to the debug output (qDebug), the construction is not aborted, though.
If copy is set to true, data points in data will only be copied. if false, the graph takes ownership of the passed data and replaces the internal data pointer with it. This is significantly faster than copying for large datasets.
void QCPGraph::setData
(
const QVector< double > &
key,
const QVector< double > &
value
)
This is an overloaded function.
Replaces the current data with the provided points in key and value pairs. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
void QCPGraph::setDataKeyError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
keyError
)
Replaces the current data with the provided points in key and value pairs. Additionally the symmetrical key error of the data points are set to the values in keyError. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
For asymmetrical errors (plus different from minus), see the overloaded version of this function.
void QCPGraph::setDataKeyError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
keyErrorMinus,
const QVector< double > &
keyErrorPlus
)
This is an overloaded function.
Replaces the current data with the provided points in key and value pairs. Additionally the negative key error of the data points are set to the values in keyErrorMinus, the positive key error to keyErrorPlus. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
void QCPGraph::setDataValueError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
valueError
)
Replaces the current data with the provided points in key and value pairs. Additionally the symmetrical value error of the data points are set to the values in valueError. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
For asymmetrical errors (plus different from minus), see the overloaded version of this function.
void QCPGraph::setDataValueError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
valueErrorMinus,
const QVector< double > &
valueErrorPlus
)
This is an overloaded function.
Replaces the current data with the provided points in key and value pairs. Additionally the negative value error of the data points are set to the values in valueErrorMinus, the positive value error to valueErrorPlus. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
void QCPGraph::setDataBothError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
keyError,
const QVector< double > &
valueError
)
Replaces the current data with the provided points in key and value pairs. Additionally the symmetrical key and value errors of the data points are set to the values in keyError and valueError. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
For asymmetrical errors (plus different from minus), see the overloaded version of this function.
void QCPGraph::setDataBothError
(
const QVector< double > &
key,
const QVector< double > &
value,
const QVector< double > &
keyErrorMinus,
const QVector< double > &
keyErrorPlus,
const QVector< double > &
valueErrorMinus,
const QVector< double > &
valueErrorPlus
)
This is an overloaded function.
Replaces the current data with the provided points in key and value pairs. Additionally the negative key and value errors of the data points are set to the values in keyErrorMinus and valueErrorMinus. The positive key and value errors are set to the values in keyErrorPlusvalueErrorPlus. For error bars to show appropriately, see setErrorType. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
Sets the visual appearance of single data points in the plot. If set to QCPScatterStyle::ssNone, no scatter points are drawn (e.g. for line-only-plots with appropriate line style).
Sets which kind of error bars (Key Error, Value Error or both) should be drawn on each data point. If you set errorType to something other than etNone, make sure to actually pass error data via the specific setData functions along with the data points (e.g. setDataValueError, setDataKeyError, setDataBothError).
Sets the width of the handles at both ends of an error bar in pixels.
void QCPGraph::setErrorBarSkipSymbol
(
bool
enabled
)
If enabled is set to true, the error bar will not be drawn as a solid line under the scatter symbol but leave some free space around the symbol.
This feature uses the current scatter size (QCPScatterStyle::setSize) to determine the size of the area to leave blank. So when drawing Pixmaps as scatter points (QCPScatterStyle::ssPixmap), the scatter size must be set manually to a value corresponding to the size of the Pixmap, if the error bars should leave gaps to its boundaries.
Removes all data points with keys between fromKey and toKey. if fromKey is greater or equal to toKey, the function does nothing. To remove a single data point with known key, use removeData(double key).
Removes a single data point at key. If the position is not known with absolute precision, consider using removeData(double fromKey, double toKey) with a small fuzziness interval around the suspected position, depeding on the precision with which the key is known.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
called by rescaleAxes functions to get the full data key bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
called by rescaleAxes functions to get the full data value bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
Draws the fill of the graph with the specified brush.
If the fill is a normal fill towards the zero-value-line, only the lineData is required (and two extra points at the zero-value-line, which are added by addFillBasePoints and removed by removeFillBasePoints after the fill drawing is done).
If the fill is a channel fill between this QCPGraph and another QCPGraph (mChannelFillGraph), the more complex polygon is calculated with the getChannelFillPolygon function.
Draws scatter symbols at every data point passed in pointData. scatter symbols are independent of the line style and are always drawn if the scatter style's shape is not QCPScatterStyle::ssNone. Hence, the pointData vector is outputted by all "get(...)PlotData" functions, together with the (line style dependent) line data.
Draws line graphs from the provided data. It connects all points in lineData, which was created by one of the "get(...)PlotData" functions for line styles that require simple line connections between the point vector they create. These are for example getLinePlotData, getStepLeftPlotData, getStepRightPlotData and getStepCenterPlotData.
This function branches out to the line style specific "get(...)PlotData" functions, according to the line style of the graph.
lineData will be filled with raw points that will be drawn with the according draw functions, e.g. drawLinePlot and drawImpulsePlot. These aren't necessarily the original data points, since for step plots for example, additional points are needed for drawing lines that make up steps. If the line style of the graph is lsNone, the lineData vector will be left untouched.
pointData will be filled with the original data points so drawScatterPlot can draw the scatter symbols accordingly. If no scatters need to be drawn, i.e. the scatter style's shape is QCPScatterStyle::ssNone, pass 0 as pointData, and this step will be skipped.
If line style is lsNone and the scatter style's shape is not QCPScatterStyle::ssNone, this function serves at providing the visible data points in pointData, so the drawScatterPlot function can draw the scatter points accordingly.
If line style is not lsNone, this function is not called and the data for the scatter points are (if needed) calculated inside the corresponding other "get(...)PlotData" functions.
Places the raw data points needed for a normal linearly connected graph in lineData.
As for all plot data retrieval functions, pointData just contains all unaltered data (scatter) points that are visible for drawing scatter points, if necessary. If drawing scatter points is disabled (i.e. the scatter style's shape is QCPScatterStyle::ssNone), pass 0 as pointData, and the function will skip filling the vector.
Places the raw data points needed for a step plot with left oriented steps in lineData.
As for all plot data retrieval functions, pointData just contains all unaltered data (scatter) points that are visible for drawing scatter points, if necessary. If drawing scatter points is disabled (i.e. the scatter style's shape is QCPScatterStyle::ssNone), pass 0 as pointData, and the function will skip filling the vector.
Places the raw data points needed for a step plot with right oriented steps in lineData.
As for all plot data retrieval functions, pointData just contains all unaltered data (scatter) points that are visible for drawing scatter points, if necessary. If drawing scatter points is disabled (i.e. the scatter style's shape is QCPScatterStyle::ssNone), pass 0 as pointData, and the function will skip filling the vector.
Places the raw data points needed for a step plot with centered steps in lineData.
As for all plot data retrieval functions, pointData just contains all unaltered data (scatter) points that are visible for drawing scatter points, if necessary. If drawing scatter points is disabled (i.e. the scatter style's shape is QCPScatterStyle::ssNone), pass 0 as pointData, and the function will skip filling the vector.
Places the raw data points needed for an impulse plot in lineData.
As for all plot data retrieval functions, pointData just contains all unaltered data (scatter) points that are visible for drawing scatter points, if necessary. If drawing scatter points is disabled (i.e. the scatter style's shape is QCPScatterStyle::ssNone), pass 0 as pointData, and the function will skip filling the vector.
called by the scatter drawing function (drawScatterPlot) to draw the error bars on one data point. x and y pixel positions of the data point are passed since they are already known in pixel coordinates in the drawing function, so we save some extra coordToPixel transforms here. data is therefore only used for the errors, not key and value.
void QCPGraph::getVisibleDataBounds
(
QCPDataMap::const_iterator &
lower,
QCPDataMap::const_iterator &
upper,
int &
count
)
const
protected
called by the specific plot data generating functions "get(...)PlotData" to determine which data range is visible, so only that needs to be processed.
lower returns an iterator to the lowest data point that needs to be taken into account when plotting. Note that in order to get a clean plot all the way to the edge of the axes, lower may still be outside the visible range.
upper returns an iterator to the highest data point. Same as before, upper may also lie outside of the visible range.
count number of data points that need plotting, i.e. points between lower and upper, including them. This is useful for allocating the array of QPointFs in the specific drawing functions.
if the graph contains no data, count is zero and both lower and upper point to constEnd.
void QCPGraph::addFillBasePoints
(
QVector< QPointF > *
lineData
)
const
protected
The line data vector generated by e.g. getLinePlotData contains only the line that connects the data points. If the graph needs to be filled, two additional points need to be added at the value-zero-line in the lower and upper key positions of the graph. This function calculates these points and adds them to the end of lineData. Since the fill is typically drawn before the line stroke, these added points need to be removed again after the fill is done, with the removeFillBasePoints function.
The expanding of lineData by two points will not cause unnecessary memory reallocations, because the data vector generation functions (getLinePlotData etc.) reserve two extra points when they allocate memory for lineData.
called by addFillBasePoints to conveniently assign the point which closes the fill polygon on the lower side of the zero-value-line parallel to the key axis. The logarithmic axis scale case is a bit special, since the zero-value-line in pixel coordinates is in positive or negative infinity. So this case is handled separately by just closing the fill polygon on the axis which lies in the direction towards the zero value.
lowerKey will be the the key (in pixels) of the returned point. Depending on whether the key axis is horizontal or vertical, lowerKey will end up as the x or y value of the returned point, respectively.
called by addFillBasePoints to conveniently assign the point which closes the fill polygon on the upper side of the zero-value-line parallel to the key axis. The logarithmic axis scale case is a bit special, since the zero-value-line in pixel coordinates is in positive or negative infinity. So this case is handled separately by just closing the fill polygon on the axis which lies in the direction towards the zero value.
upperKey will be the the key (in pixels) of the returned point. Depending on whether the key axis is horizontal or vertical, upperKey will end up as the x or y value of the returned point, respectively.
Generates the polygon needed for drawing channel fills between this graph (data passed via lineData) and the graph specified by mChannelFillGraph (data generated by calling its getPlotData function). May return an empty polygon if the key ranges have no overlap or fill target graph and this graph don't have same orientation (i.e. both key axes horizontal or both key axes vertical). For increased performance (due to implicit sharing), keep the returned QPolygonF const.
int QCPGraph::findIndexBelowX
(
const QVector< QPointF > *
data,
double
x
)
const
protected
Finds the highest index of data, whose points x value is just below x. Assumes x values in data points are ordered ascending, as is the case when plotting with horizontal key axis.
Finds the smallest index of data, whose points x value is just above x. Assumes x values in data points are ordered ascending, as is the case when plotting with horizontal key axis.
Finds the highest index of data, whose points y value is just below y. Assumes y values in data points are ordered descending, as is the case when plotting with vertical key axis (since keys are ordered ascending).
Finds the smallest index of data, whose points y value is just above y. Assumes y values in data points are ordered descending, as is the case when plotting with vertical key axis.
Calculates the (minimum) distance (in pixels) the graph's representation has from the given pixelPoint in pixels. This is used to determine whether the graph was clicked or not, e.g. in selectTest.
If either the graph has no data or if the line style is lsNone and the scatter style's shape is QCPScatterStyle::ssNone (i.e. there is no visual representation of the graph), returns 500.
void QCPAbstractPlottable::setName
(
const QString &
name
)
inherited
The name is the textual representation of this plottable as it is displayed in the legend (QCPLegend). It may contain any UTF-8 characters, including newlines.
void QCPAbstractPlottable::setAntialiasedFill
(
bool
enabled
)
inherited
Sets whether fills of this plottable is drawn antialiased or not.
The brush is used to draw basic fills of the plottable representation in the plot. The Fill can be a color, gradient or texture, see the usage of QBrush.
For example, the QCPGraph subclass draws the fill under the graph with this brush, when it's not set to Qt::NoBrush.
The key axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's value axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
The value axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's key axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
Sets whether the user can (de-)select this plottable by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains iSelectPlottables.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected directly.
Sets whether this plottable is selected or not. When selected, it uses a different pen and brush to draw its lines and fills, see setSelectedPen and setSelectedBrush.
The entire selection mechanism for plottables is handled automatically when QCustomPlot::setInteractions contains iSelectPlottables. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Adds this plottable to the legend of the parent QCustomPlot (QCustomPlot::legend).
Normally, a QCPPlottableLegendItem is created and inserted into the legend. If the plottable needs a more specialized representation in the legend, this function will take this into account and instead create the specialized subclass of QCPAbstractLegendItem.
Returns true on success, i.e. when the legend exists and a legend item associated with this plottable isn't already in the legend.
Rescales the key and value axes associated with this plottable to contain all displayed data, so the whole plottable is visible. If the scaling of an axis is logarithmic, rescaleAxes will make sure not to rescale to an illegal range i.e. a range containing different signs and/or zero. Instead it will stay in the current sign domain and ignore all parts of the plottable that lie outside of that domain.
onlyEnlarge makes sure the ranges are only expanded, never reduced. So it's possible to show multiple plottables in their entirety by multiple calls to rescaleAxes where the first call has onlyEnlarge set to false (the default), and all subsequent set to true.
This signal is emitted when the selection state of this plottable has changed to selected, either by user interaction or by a direct call to setSelected.
QRect QCPAbstractPlottable::clipRect
(
)
const
protectedvirtualinherited
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Convenience function for transforming a key/value pair to pixels on the QCustomPlot surface, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
key and value are transformed to the coodinates in pixels and are written to x and y.
Returns the input as pixel coordinates in a QPointF.
void QCPAbstractPlottable::pixelsToCoords
(
double
x,
double
y,
double &
key,
double &
value
)
const
protectedinherited
Convenience function for transforming a x/y pixel pair on the QCustomPlot surface to plot coordinates, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
x and y are transformed to the plot coodinates and are written to key and value.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
Returns the pen that should be used for drawing lines. Returns mPen when the item is not selected and mSelectedPen when it is.
QBrush QCPItemRect::mainBrush
(
)
const
protected
Returns the brush that should be used for drawing fills of the item. Returns mBrush when the item is not selected and mSelectedBrush when it is.
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
inherited
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
To plot data, assign it with the setData or addData functions.
Changing the appearance
The appearance of the bars is determined by the pen and the brush (setPen, setBrush).
Bar charts are stackable. This means, Two QCPBars plottables can be placed on top of each other (see QCPBars::moveAbove). Then, when two bars are at the same key position, they will appear stacked.
Constructs a bar chart which uses keyAxis as its key axis ("x") and valueAxis as its value axis ("y"). keyAxis and valueAxis must reside in the same QCustomPlot instance and not have the same orientation. If either of these restrictions is violated, a corresponding message is printed to the debug output (qDebug), the construction is not aborted, though.
If copy is set to true, data points in data will only be copied. if false, the plottable takes ownership of the passed data and replaces the internal data pointer with it. This is significantly faster than copying for large datasets.
void QCPBars::setData
(
const QVector< double > &
key,
const QVector< double > &
value
)
This is an overloaded function.
Replaces the current data with the provided points in key and value tuples. The provided vectors should have equal length. Else, the number of added points will be the size of the smallest vector.
Moves this bars plottable below bars. In other words, the bars of this plottable will appear below the bars of bars. The move target bars must use the same key and value axis as this plottable.
Inserting into and removing from existing bar stacking is handled gracefully. If bars already has a bars object below itself, this bars object is inserted between the two. If this bars object is already between two other bars, the two other bars will be stacked on top of each other after the operation.
To remove this bars plottable from any stacking, set bars to 0.
Moves this bars plottable above bars. In other words, the bars of this plottable will appear above the bars of bars. The move target bars must use the same key and value axis as this plottable.
Inserting into and removing from existing bar stacking is handled gracefully. If bars already has a bars object below itself, this bars object is inserted between the two. If this bars object is already between two other bars, the two other bars will be stacked on top of each other after the operation.
To remove this bars plottable from any stacking, set bars to 0.
Removes all data points with key between fromKey and toKey. if fromKey is greater or equal to toKey, the function does nothing. To remove a single data point with known key, use removeData(double key).
Removes a single data point at key. If the position is not known with absolute precision, consider using removeData(double fromKey, double toKey) with a small fuzziness interval around the suspected position, depeding on the precision with which the key is known.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
This function draws the layerable with the specified painter. It is only called by QCustomPlot, if the layerable is visible (setVisible).
Before this function is called, the painter's antialiasing state is set via applyDefaultAntialiasingHint, see the documentation there. Further, the clipping rectangle was set to clipRect.
called by rescaleAxes functions to get the full data key bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
called by rescaleAxes functions to get the full data value bounds. For logarithmic plots, one can set inSignDomain to either sdNegative or sdPositive in order to restrict the returned range to that sign domain. E.g. when only negative range is wanted, set inSignDomain to sdNegative and all positive points will be ignored for range calculation. For no restriction, just set inSignDomain to sdBoth (default). validRange is an output parameter that indicates whether a proper range could be found or not. If this is false, you shouldn't use the returned range (e.g. no points in data).
Returns the polygon of a single bar with key and value. The Polygon is open at the bottom and shifted according to the bar stacking (see moveAbove).
double QCPBars::getBaseValue
(
double
key,
bool
positive
)
const
protected
This function is called to find at which value to start drawing the base of a bar at key, when it is stacked on top of another QCPBars (e.g. with moveAbove).
positive and negative bars are separated per stack (positive are stacked above 0-value upwards, negative are stacked below 0-value downwards). This can be indicated with positive. So if the bar for which we need the base value is negative, set positive to false.
Connects below and above to each other via their mBarAbove/mBarBelow properties. The bar(s) currently below lower and upper will become disconnected to lower/upper.
If lower is zero, upper will be disconnected at the bottom. If upper is zero, lower will be disconnected at the top.
void QCPAbstractPlottable::setName
(
const QString &
name
)
inherited
The name is the textual representation of this plottable as it is displayed in the legend (QCPLegend). It may contain any UTF-8 characters, including newlines.
void QCPAbstractPlottable::setAntialiasedFill
(
bool
enabled
)
inherited
Sets whether fills of this plottable is drawn antialiased or not.
The brush is used to draw basic fills of the plottable representation in the plot. The Fill can be a color, gradient or texture, see the usage of QBrush.
For example, the QCPGraph subclass draws the fill under the graph with this brush, when it's not set to Qt::NoBrush.
The key axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's value axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
The value axis of a plottable can be set to any axis of a QCustomPlot, as long as it is orthogonal to the plottable's key axis. This function performs no checks to make sure this is the case. The typical mathematical choice is to use the x-axis (QCustomPlot::xAxis) as key axis and the y-axis (QCustomPlot::yAxis) as value axis.
Normally, the key and value axes are set in the constructor of the plottable (or QCustomPlot::addGraph when working with QCPGraphs through the dedicated graph interface).
Sets whether the user can (de-)select this plottable by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains iSelectPlottables.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected directly.
Sets whether this plottable is selected or not. When selected, it uses a different pen and brush to draw its lines and fills, see setSelectedPen and setSelectedBrush.
The entire selection mechanism for plottables is handled automatically when QCustomPlot::setInteractions contains iSelectPlottables. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Adds this plottable to the legend of the parent QCustomPlot (QCustomPlot::legend).
Normally, a QCPPlottableLegendItem is created and inserted into the legend. If the plottable needs a more specialized representation in the legend, this function will take this into account and instead create the specialized subclass of QCPAbstractLegendItem.
Returns true on success, i.e. when the legend exists and a legend item associated with this plottable isn't already in the legend.
Rescales the key and value axes associated with this plottable to contain all displayed data, so the whole plottable is visible. If the scaling of an axis is logarithmic, rescaleAxes will make sure not to rescale to an illegal range i.e. a range containing different signs and/or zero. Instead it will stay in the current sign domain and ignore all parts of the plottable that lie outside of that domain.
onlyEnlarge makes sure the ranges are only expanded, never reduced. So it's possible to show multiple plottables in their entirety by multiple calls to rescaleAxes where the first call has onlyEnlarge set to false (the default), and all subsequent set to true.
This signal is emitted when the selection state of this plottable has changed to selected, either by user interaction or by a direct call to setSelected.
QRect QCPAbstractPlottable::clipRect
(
)
const
protectedvirtualinherited
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Convenience function for transforming a key/value pair to pixels on the QCustomPlot surface, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
key and value are transformed to the coodinates in pixels and are written to x and y.
Returns the input as pixel coordinates in a QPointF.
void QCPAbstractPlottable::pixelsToCoords
(
double
x,
double
y,
double &
key,
double &
value
)
const
protectedinherited
Convenience function for transforming a x/y pixel pair on the QCustomPlot surface to plot coordinates, taking the orientations of the axes associated with this plottable into account (e.g. whether key represents x or y).
x and y are transformed to the plot coodinates and are written to key and value.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Pixmap example. Blue dotted circles are anchors, solid blue discs are positions.
It has two positions, topLeft and bottomRight, which define the rectangle the pixmap will be drawn in. Depending on the scale setting (setScaled), the pixmap will be either scaled to fit the rectangle or be drawn aligned to the topLeft position.
If scaling is enabled and topLeft is further to the bottom/right than bottomRight (as shown on the right side of the example image), the pixmap will be flipped in the respective orientations.
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
Creates the buffered scaled image (mScaledPixmap) to fit the specified finalRect. The parameters flipHorz and flipVert control whether the resulting image shall be flipped horizontally or vertically. (This is used when topLeft is further to the bottom/right than bottomRight.)
This function only creates the scaled pixmap when the buffered pixmap has a different size than the expected result, so calling this function repeatedly, e.g. in the draw function, does not cause expensive rescaling every time.
If scaling is disabled, sets mScaledPixmap to a null QPixmap.
QRect QCPItemPixmap::getFinalRect
(
bool *
flippedHorz = 0,
bool *
flippedVert = 0
)
const
protected
Returns the final (tight) rect the pixmap is drawn in, depending on the current item positions and scaling settings.
The output parameters flippedHorz and flippedVert return whether the pixmap should be drawn flipped horizontally or vertically in the returned rect. (The returned rect itself is always normalized, i.e. the top left corner of the rect is actually further to the top/left than the bottom right corner). This is the case when the item position topLeft is further to the bottom/right than bottomRight.
If scaling is disabled, returns a rect with size of the original pixmap and the top left corner aligned with the item position topLeft. The position bottomRight is ignored.
QPen QCPItemPixmap::mainPen
(
)
const
protected
Returns the pen that should be used for drawing lines. Returns mPen when the item is not selected and mSelectedPen when it is.
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
inherited
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
The QCPLegend derives from QCPLayoutGrid and as such can be placed in any position a QCPLayoutElement may be positioned. The legend items are themselves QCPLayoutElements which are placed in the grid layout of the legend. QCPLegend only adds an interface specialized for handling child elements of type QCPAbstractLegendItem, as mentioned above. In principle, any other layout elements may also be added to a legend via the normal QCPLayoutGrid interface. However, the QCPAbstractLegendItem-Interface will ignore those elements (e.g. itemCount will only return the number of items with QCPAbstractLegendItems type).
By default, every QCustomPlot has one legend (QCustomPlot::legend) which is placed in the inset layout of the main axis rect (QCPAxisRect::insetLayout). To move the legend to another position inside the axis rect, use the methods of the QCPLayoutInset. To move the legend outside of the axis rect, place it anywhere else with the QCPLayout/QCPLayoutElement interface.
Constructs a new QCPLegend instance with parentPlot as the containing plot and default values.
Note that by default, QCustomPlot already contains a legend ready to be used as QCustomPlot::legend
Member Function Documentation
void QCPLegend::setBorderPen
(
const QPen &
pen
)
Sets the pen, the border of the entire legend is drawn with.
void QCPLegend::setBrush
(
const QBrush &
brush
)
Sets the brush of the legend background.
void QCPLegend::setFont
(
const QFont &
font
)
Sets the default font of legend text. Legend items that draw text (e.g. the name of a graph) will use this font by default. However, a different font can be specified on a per-item-basis by accessing the specific legend item.
This function will also set font on all already existing legend items.
Sets the default color of legend text. Legend items that draw text (e.g. the name of a graph) will use this color by default. However, a different colors can be specified on a per-item-basis by accessing the specific legend item.
This function will also set color on all already existing legend items.
Sets the size of legend icons. Legend items that draw an icon (e.g. a visual representation of the graph) will use this size by default.
void QCPLegend::setIconSize
(
int
width,
int
height
)
This is an overloaded function.
void QCPLegend::setIconTextPadding
(
int
padding
)
Sets the horizontal space in pixels between the legend icon and the text next to it. Legend items that draw an icon (e.g. a visual representation of the graph) and text (e.g. the name of the graph) will use this space by default.
void QCPLegend::setIconBorderPen
(
const QPen &
pen
)
Sets the pen used to draw a border around each legend icon. Legend items that draw an icon (e.g. a visual representation of the graph) will use this pen by default.
If no border is wanted, set this to Qt::NoPen.
void QCPLegend::setSelectableParts
(
const SelectableParts &
selectable
)
Sets whether the user can (de-)select the parts in selectable by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains iSelectLegend.)
However, even when selectable is set to a value not allowing the selection of a specific part, it is still possible to set the selection of this part manually, by calling setSelectedParts directly.
Sets the selected state of the respective legend parts described by SelectablePart. When a part is selected, it uses a different pen/font and brush. If some legend items are selected and selected doesn't contain spItems, those items become deselected.
The entire selection mechanism is handled automatically when QCustomPlot::setInteractions contains iSelectLegend. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state of a part even when setSelectableParts was set to a value that actually excludes the part.
emits the selectionChanged signal when selected is different from the previous selection state.
Note that it doesn't make sense to set the selected state spItems here when it wasn't set before, because there's no way to specify which exact items to newly select. Do this by calling QCPAbstractLegendItem::setSelected directly on the legend item you wish to select.
Layout elements are sensitive to events inside their outer rect. If pos is within the outer rect, this method returns a value corresponding to 0.99 times the parent plot's selection tolerance. However, layout elements are not selectable by default. So if onlySelectable is true, -1.0 is returned.
Returns whether the legend contains a QCPPlottableLegendItem which is associated with plottable (e.g. a QCPGraph*). If such an item isn't in the legend, returns false.
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
Subclasses reimplement this method to update the position and sizes of the child elements/cells via calling their QCPLayoutElement::setOuterRect. The default implementation does nothing.
The geometry used as a reference is the inner rect of this layout. Child elements should stay within that rect.
getSectionSizes may help with the reimplementation of this function.
Returns the element in the cell with the given index. If index is invalid, returns 0.
Note that even if index is valid, the respective cell may be empty in some layouts (e.g. QCPLayoutGrid), so this function may return 0 in those cases. You may use this function to check whether a cell is empty or not.
Removes the element with the given index from the layout and returns it.
If the index is invalid or the cell with that index is empty, returns 0.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Removes the specified element from the layout and returns true on success.
If the element isn't in this layout, returns false.
Note that some layouts don't remove the respective cell right away but leave an empty cell after successful removal of the layout element. To collapse empty cells, use simplify.
Returns a list of all child elements in this layout element. If recursive is true, all sub-child elements are included in the list, too.
Note that there may be entries with value 0 in the returned list. (For example, QCPLayoutGrid may have empty cells which yield 0 at the respective index.)
Returns the minimum size this layout element (the inner rect) may be compressed to.
if a minimum size (setMinimumSize) was not set manually, parent layouts consult this function to determine the minimum allowed size of this layout element. (A manual minimum size is considered set if it is non-zero.)
Returns the maximum size this layout element (the inner rect) may be expanded to.
if a maximum size (setMaximumSize) was not set manually, parent layouts consult this function to determine the maximum allowed size of this layout element. (A manual maximum size is considered set if it is smaller than Qt's QWIDGETSIZE_MAX.)
Returns the element in the cell in row and column.
Returns 0 if either the row/column is invalid or if the cell is empty. In those cases, a qDebug message is printed. To check whether a cell exists and isn't empty, use hasElement.
Adds the element to cell with row and column. If element is already in a layout, it is first removed from there. If row or column don't exist yet, the layout is expanded accordingly.
Returns true if the element was added successfully, i.e. if the cell at row and column didn't already have an element.
Expands the layout to have newRowCount rows and newColumnCount columns. So the last valid row index will be newRowCount-1, the last valid column index will be newColumnCount-1.
If the current column/row count is already larger or equal to newColumnCount/newRowCount, this function does nothing in that dimension.
Newly created cells are empty, new rows and columns have the stretch factor 1.
Note that upon a call to addElement, the layout is expanded automatically to contain the specified row and column, using this function.
Inserts a new row with empty cells at the row index newIndex. Valid values for newIndex range from 0 (inserts a row at the top) to rowCount (appends a row at the bottom).
Inserts a new column with empty cells at the column index newIndex. Valid values for newIndex range from 0 (inserts a row at the left) to rowCount (appends a row at the right).
Places the minimum column widths and row heights into minColWidths and minRowHeights respectively.
The minimum height of a row is the largest minimum height of any element in that row. The minimum width of a column is the largest minimum width of any element in that column.
Places the maximum column widths and row heights into maxColWidths and maxRowHeights respectively.
The maximum height of a row is the smallest maximum height of any element in that row. The maximum width of a column is the smallest maximum width of any element in that column.
Subclasses call this method to report changed (minimum/maximum) size constraints.
If the parent of this layout is again a QCPLayout, forwards the call to the parent's sizeConstraintsChanged. If the parent is a QWidget (i.e. is the QCustomPlot::plotLayout of QCustomPlot), calls QWidget::updateGeometry, so if the QCustomPlot widget is inside a Qt QLayout, it may update itself and resize cells accordingly.
This method is used by subclass specific methods that add elements to the layout. Note that this method only changes properties in el. The removal from the old layout and the insertion into the new layout must be done additionally.
This method is used by subclass specific methods that remove elements from the layout (e.g. take or takeAt). Note that this method only changes properties in el. The removal from the old layout must be done additionally.
QVector< int > QCPLayout::getSectionSizes
(
QVector< int >
maxSizes,
QVector< int >
minSizes,
QVector< double >
stretchFactors,
int
totalSize
)
const
protectedinherited
This is a helper function for the implementation of updateLayout in subclasses.
It calculates the sizes of one-dimensional sections with provided constraints on maximum section sizes, minimum section sizes, relative stretch factors and the final total size of all sections.
The QVector entries refer to the sections. Thus all QVectors must have the same size.
maxSizes gives the maximum allowed size of each section. If there shall be no maximum size imposed, set all vector values to Qt's QWIDGETSIZE_MAX.
minSizes gives the minimum allowed size of each section. If there shall be no minimum size imposed, set all vector values to zero. If the minSizes entries add up to a value greater than totalSize, sections will be scaled smaller than the proposed minimum sizes. (In other words, not exceeding the allowed total size is taken to be more important than not going below minimum section sizes.)
stretchFactors give the relative proportions of the sections to each other. If all sections shall be scaled equally, set all values equal. If the first section shall be double the size of each individual other section, set the first number of stretchFactors to double the value of the other individual values (e.g. {2, 1, 1, 1}).
totalSize is the value that the final section sizes will add up to. Due to rounding, the actual sum may differ slightly. If you want the section sizes to sum up to exactly that value, you could distribute the remaining difference on the sections.
The return value is a QVector containing the section sizes.
Returns the inner rect of this layout element. The inner rect is the outer rect (setOuterRect) shrinked by the margins (setMargins, setAutoMargins).
In some cases, the area between outer and inner rect is left blank. In other cases the margin area is used to display peripheral graphics while the main content is in the inner rect. This is where automatic margin calculation becomes interesting because it allows the layout element to adapt the margins to the peripheral graphics it wants to draw. For example, QCPAxisRect draws the axis labels and tick labels in the margin area, thus needs to adjust the margins (if setAutoMargins is enabled) according to the space required by the labels of the axes.
void QCPLayoutElement::setOuterRect
(
const QRect &
rect
)
inherited
Sets the outer rect of this layout element. If the layout element is inside a layout, the layout sets the position and size of this layout element using this function.
Calling this function externally has no effect, since the layout will overwrite any changes to the outer rect upon the next replot.
The layout element will adapt its inner rect by applying the margins inward to the outer rect.
Sets the margins of this layout element. If setAutoMargins is disabled for some or all sides, this function is used to manually set the margin on those sides. Sides that are still set to be handled automatically are ignored and may have any value in margins.
The margin is the distance between the outer rect (controlled by the parent layout via setOuterRect) and the inner rect (which usually contains the main content of this layout element).
Sets on which sides the margin shall be calculated automatically. If a side is calculated automatically, a minimum margin value may be provided with setMinimumMargins. If a side is set to be controlled manually, the value may be specified with setMargins.
Margin sides that are under automatic control may participate in a QCPMarginGroup (see setMarginGroup), to synchronize (align) it with other layout elements in the plot.
Sets the minimum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
If the parent layout size is not sufficient to satisfy all minimum size constraints of its child layout elements, the layout may set a size that is actually smaller than size. QCustomPlot propagates the layout's size constraints to the outside by setting its own minimum QWidget size accordingly, so violations of size should be exceptions.
void QCPLayoutElement::setMinimumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the minimum size for the inner rect of this layout element.
void QCPLayoutElement::setMaximumSize
(
const QSize &
size
)
inherited
Sets the maximum size for the inner rect of this layout element. A parent layout tries to respect the size here by changing row/column sizes in the layout accordingly.
void QCPLayoutElement::setMaximumSize
(
int
width,
int
height
)
inherited
This is an overloaded function.
Sets the maximum size for the inner rect of this layout element.
Returns the margin size for this side. It is used if automatic margins is enabled for this side (see setAutoMargins). If a minimum margin was set with setMinimumMargins, the returned value will not be smaller than the specified minimum margin.
The default implementation just returns the respective manual margin (setMargins) or the minimum margin, whichever is larger.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
QRect QCPLayerable::clipRect
(
)
const
protectedvirtualinherited
Returns the clipping rectangle of this layerable object. By default, this is the viewport of the parent QCustomPlot. Specific subclasses may reimplement this function to provide different clipping rects.
The returned clipping rect is set on the painter before the draw function of the respective object is called.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
This internal class is used to provide some extended functionality e.g. for tweaking position consistency between antialiased and non-antialiased painting. Further it provides workarounds for QPainter quirks.
Warning
This class intentionally hides non-virtual functions of QPainter, e.g. setPen, save and restore. So while it is possible to pass a QCPPainter instance to a function that expects a QPainter pointer, some of the workarounds and tweaks will be unavailable to the function (because it will call the base class implementations of the functions actually hidden by QCPPainter).
Defines special modes the painter can operate in. They disable or enable certain subsets of features/fixes/workarounds, depending on whether they are wanted on the respective output device.
Enumerator:
pmDefault
0x00 Default mode for painting on screen devices
pmVectorized
0x01 Mode for vectorized painting (e.g. PDF export). For example, this prevents some antialiasing fixes.
pmNoCaching
0x02 Mode for all sorts of exports (e.g. PNG, PDF,...). For example, this prevents using cached pixmap labels
pmNonCosmetic
0x04 Turns pen widths 0 to 1, i.e. disables cosmetic pens. (A cosmetic pen is always drawn with width 1 pixel in the vector image/pdf viewer, independent of zoom.)
Constructor & Destructor Documentation
QCPPainter::QCPPainter
(
)
Creates a new QCPPainter instance and sets default values
QCPPainter::QCPPainter
(
QPaintDevice *
device
)
Creates a new QCPPainter instance on the specified paint device and sets default values. Just like the analogous QPainter constructor, begins painting on device immediately.
Like begin, this method sets QPainter::NonCosmeticDefaultPen in Qt versions before Qt5.
Member Function Documentation
void QCPPainter::setAntialiasing
(
bool
enabled
)
Sets whether painting uses antialiasing or not. Use this method instead of using setRenderHint with QPainter::Antialiasing directly, as it allows QCPPainter to regain pixel exactness between antialiased and non-antialiased painting (Since Qt < 5.0 uses slightly different coordinate systems for AA/Non-AA painting).
Sets the mode of the painter. This controls whether the painter shall adjust its fixes/workarounds optimized for certain output devices.
void QCPPainter::setModes
(
PainterModes
modes
)
Sets the mode of the painter. This controls whether the painter shall adjust its fixes/workarounds optimized for certain output devices.
bool QCPPainter::begin
(
QPaintDevice *
device
)
Sets the QPainter::NonCosmeticDefaultPen in Qt versions before Qt5 after beginning painting on device. This is necessary to get cosmetic pen consistency across Qt versions, because since Qt5, all pens are non-cosmetic by default, and in Qt4 this render hint must be set to get that behaviour.
this function hides the non-virtual base class implementation.
void QCPPainter::setPen
(
const QPen &
pen
)
Sets the pen of the painter and applies certain fixes to it, depending on the mode of this QCPPainter.
Note
this function hides the non-virtual base class implementation.
void QCPPainter::setPen
(
const QColor &
color
)
This is an overloaded function.
Sets the pen (by color) of the painter and applies certain fixes to it, depending on the mode of this QCPPainter.
Note
this function hides the non-virtual base class implementation.
void QCPPainter::setPen
(
Qt::PenStyle
penStyle
)
This is an overloaded function.
Sets the pen (by style) of the painter and applies certain fixes to it, depending on the mode of this QCPPainter.
Note
this function hides the non-virtual base class implementation.
void QCPPainter::drawLine
(
const QLineF &
line
)
This is an overloaded function.
Works around a Qt bug introduced with Qt 4.8 which makes drawing QLineF unpredictable when antialiasing is disabled. Thus when antialiasing is disabled, it rounds the line to integer coordinates and then passes it to the original drawLine.
Note
this function hides the non-virtual base class implementation.
void QCPPainter::save
(
)
Saves the painter (see QPainter::save). Since QCPPainter adds some new internal state to QPainter, the save/restore functions are reimplemented to also save/restore those members.
Note
this function hides the non-virtual base class implementation.
Restores the painter (see QPainter::restore). Since QCPPainter adds some new internal state to QPainter, the save/restore functions are reimplemented to also save/restore those members.
Note
this function hides the non-virtual base class implementation.
Text example. Blue dotted circles are anchors, solid blue discs are positions.
Its position is defined by the member position and the setting of setPositionAlignment. The latter controls which part of the text rect shall be aligned with position.
The text alignment itself (i.e. left, center, right) can be controlled with setTextAlignment.
The text may be rotated around the position point with setRotation.
Sets which point of the text rect shall be aligned with position.
Examples:
If alignment is Qt::AlignHCenter | Qt::AlignTop, the text will be positioned such that the top of the text rect will be horizontally centered on position.
If alignment is Qt::AlignLeft | Qt::AlignBottom, position will indicate the bottom left corner of the text rect.
If you want to control the alignment of (multi-lined) text within the text rect, use setTextAlignment.
void QCPItemText::setTextAlignment
(
Qt::Alignment
alignment
)
Controls how (multi-lined) text is aligned inside the text rect (typically Qt::AlignLeft, Qt::AlignCenter or Qt::AlignRight).
void QCPItemText::setRotation
(
double
degrees
)
Sets the angle in degrees by which the text (and the text rectangle, if visible) will be rotated around position.
void QCPItemText::setPadding
(
const QMargins &
padding
)
Sets the distance between the border of the text rectangle and the text. The appearance (and visibility) of the text rectangle can be controlled with setPen and setBrush.
double QCPItemText::selectTest
(
const QPointF &
pos,
bool
onlySelectable,
QVariant *
details = 0
)
const
virtual
This function is used to decide whether a click hits a layerable object or not.
pos is a point in pixel coordinates on the QCustomPlot surface. This function returns the shortest pixel distance of this point to the object. If the object is either invisible or the distance couldn't be determined, -1.0 is returned. Further, if onlySelectable is true and the object is not selectable, -1.0 is returned, too.
If the item is represented not by single lines but by an area like QCPItemRect or QCPItemText, a click inside the area returns a constant value greater zero (typically the selectionTolerance of the parent QCustomPlot multiplied by 0.99). If the click lies outside the area, this function returns -1.0.
Providing a constant value for area objects allows selecting line objects even when they are obscured by such area objects, by clicking close to the lines (i.e. closer than 0.99*selectionTolerance).
The actual setting of the selection state is not done by this function. This is handled by the parent QCustomPlot when the mouseReleaseEvent occurs, and the finally selected object is notified via the selectEvent/deselectEvent methods.
details is an optional output parameter. Every layerable subclass may place any information in details. This information will be passed to selectEvent when the parent QCustomPlot decides on the basis of this selectTest call, that the object was successfully selected. The subsequent call to selectEvent will carry the details. This is useful for multi-part objects (like QCPAxis). This way, a possibly complex calculation to decide which part was clicked is only done once in selectTest. The result (i.e. the actually clicked part) can then be placed in details. So in the subsequent selectEvent, the decision which part was selected doesn't have to be done a second time for a single selection operation.
You may pass 0 as details to indicate that you are not interested in those selection details.
The cliprect of the provided painter is set to the rect returned by clipRect before this function is called. The clipRect depends on the clipping settings defined by setClipToAxisRect and setClipAxisRect.
Returns the pixel position of the anchor with Id anchorId. This function must be reimplemented in item subclasses if they want to provide anchors (QCPItemAnchor).
For example, if the item has two anchors with id 0 and 1, this function takes one of these anchor ids and returns the respective pixel points of the specified anchor.
Returns the point that must be given to the QPainter::drawText function (which expects the top left point of the text rect), according to the position pos, the text bounding box rect and the requested positionAlignment.
For example, if positionAlignment is Qt::AlignLeft | Qt::AlignBottom the returned point will be shifted upward by the height of rect, starting from pos. So if the text is finally drawn at that point, the lower left corner of the resulting text rect is at pos.
QFont QCPItemText::mainFont
(
)
const
protected
Returns the font that should be used for drawing text. Returns mFont when the item is not selected and mSelectedFont when it is.
QColor QCPItemText::mainColor
(
)
const
protected
Returns the color that should be used for drawing text. Returns mColor when the item is not selected and mSelectedColor when it is.
QPen QCPItemText::mainPen
(
)
const
protected
Returns the pen that should be used for drawing lines. Returns mPen when the item is not selected and mSelectedPen when it is.
QBrush QCPItemText::mainBrush
(
)
const
protected
Returns the brush that should be used for drawing fills of the item. Returns mBrush when the item is not selected and mSelectedBrush when it is.
void QCPAbstractItem::setClipToAxisRect
(
bool
clip
)
inherited
Sets whether the item shall be clipped to an axis rect or whether it shall be visible on the entire QCustomPlot. The axis rect can be set with setClipAxisRect.
Sets whether the user can (de-)select this item by clicking on the QCustomPlot surface. (When QCustomPlot::setInteractions contains QCustomPlot::iSelectItems.)
However, even when selectable was set to false, it is possible to set the selection manually, by calling setSelected.
Sets whether this item is selected or not. When selected, it might use a different visual appearance (e.g. pen and brush), this depends on the specific item though.
The entire selection mechanism for items is handled automatically when QCustomPlot::setInteractions contains QCustomPlot::iSelectItems. You only need to call this function when you wish to change the selection state manually.
This function can change the selection state even when setSelectable was set to false.
emits the selectionChanged signal when selected is different from the previous selection state.
Returns all anchors of the item in a list. Note that since a position (QCPItemPosition) is always also an anchor, the list will also contain the positions of this item.
Returns the QCPItemPosition with the specified name. If this item doesn't have a position by that name, returns 0.
This function provides an alternative way to access item positions. Normally, you access positions direcly by their member pointers (which typically have the same variable name as name).
Returns the QCPItemAnchor with the specified name. If this item doesn't have an anchor by that name, returns 0.
This function provides an alternative way to access item anchors. Normally, you access anchors direcly by their member pointers (which typically have the same variable name as name).
Returns whether this item has an anchor with the specified name.
Note that you can check for positions with this function, too. This is because every position is also an anchor (QCPItemPosition inherits from QCPItemAnchor).
Returns the selection category this layerable shall belong to. The selection category is used in conjunction with QCustomPlot::setInteractions to control which objects are selectable and which aren't.
Subclasses that don't fit any of the normal QCP::Interaction values can use QCP::iSelectOther. This is what the default implementation returns.
Returns the rect the visual representation of this item is clipped to. This depends on the current setting of setClipToAxisRect as well as the axis rect set with setClipAxisRect.
If the item is not clipped to an axis rect, the QCustomPlot::viewport rect is returned.
This event is called when the layerable shall be selected, as a consequence of a click by the user. Subclasses should react to it by setting their selection state appropriately. The default implementation does nothing.
event is the mouse event that caused the selection. additive indicates, whether the user was holding the multi-select-modifier while performing the selection (see QCustomPlot::setMultiSelectModifier). if additive is true, the selection state must be toggled (i.e. become selected when unselected and unselected when selected).
Every selectEvent is preceded by a call to selectTest, which has returned positively (i.e. returned a value greater than 0 and less than the selection tolerance of the parent QCustomPlot). The details data you output from selectTest is feeded back via details here. You may use it to transport any kind of information from the selectTest to the possibly subsequent selectEvent. Usually details is used to transfer which part was clicked, if it is a layerable that has multiple individually selectable parts (like QCPAxis). This way selectEvent doesn't need to do the calculation again to find out which part was actually clicked.
selectionStateChanged is an output parameter. If the pointer is non-null, this function must set the value either to true or false, depending on whether the selection state of this layerable was actually changed. For layerables that only are selectable as a whole and not in parts, this is simple: if additive is true, selectionStateChanged must also be set to true, because the selection toggles. If additive is false, selectionStateChanged is only set to true, if the layerable was previously unselected and now is switched to the selected state.
This event is called when the layerable shall be deselected, either as consequence of a user interaction or a call to QCustomPlot::deselectAll. Subclasses should react to it by unsetting their selection appropriately.
just as in selectEvent, the output parameter selectionStateChanged (if non-null), must return true or false when the selection state of this layerable has changed or not changed, respectively.
A convenience function which returns the selectTest value for a specified rect and a specified click position pos. filledRect defines whether a click inside the rect should also be considered a hit or whether only the rect border is sensitive to hits.
This function may be used to help with the implementation of the selectTest function for specific items.
For example, if your item consists of four rects, call this function four times, once for each rect, in your selectTest reimplementation. Finally, return the minimum of all four returned values which were greater or equal to zero. (Because this function may return -1.0 when pos doesn't hit rect at all). If all calls returned -1.0, return -1.0, too, because your item wasn't hit.
Creates a QCPItemPosition, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the position member (This is needed to provide the name-based position access to positions).
Don't delete positions created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each position member. Don't create QCPItemPositions with new yourself, because they won't be registered with the item properly.
Creates a QCPItemAnchor, registers it with this item and returns a pointer to it. The specified name must be a unique string that is usually identical to the variable name of the anchor member (This is needed to provide the name based anchor access to anchors).
The anchorId must be a number identifying the created anchor. It is recommended to create an enum (e.g. "AnchorIndex") for this on each item that uses anchors. This id is used by the anchor to identify itself when it calls QCPAbstractItem::anchorPixelPoint. That function then returns the correct pixel coordinates for the passed anchor id.
Don't delete anchors created by this function manually, as the item will take care of it.
Use this function in the constructor (initialization list) of the specific item subclass to create each anchor member. Don't create QCPItemAnchors with new yourself, because then they won't be registered with the item properly.
Returns the parent layerable of this layerable. The parent layerable is used to provide visibility hierarchies in conjunction with the method realVisibility. This way, layerables only get drawn if their parent layerables are visible, too.
Note that a parent layerable is not necessarily also the QObject parent for memory management. Further, a layerable doesn't always have a parent layerable, so this function may return 0.
A parent layerable is set implicitly with when placed inside layout elements and doesn't need to be set manually by the user.
void QCPLayerable::setVisible
(
bool
on
)
inherited
Sets the visibility of this layerable object. If an object is not visible, it will not be drawn on the QCustomPlot surface, and user interaction with it (e.g. click and selection) is not possible.
Returns whether this layerable is visible, taking possible direct layerable parent visibility into account. This is the method that is consulted to decide whether a layerable shall be drawn or not.
If this layerable has a direct layerable parent (usually set via hierarchies implemented in subclasses, like in the case of QCPLayoutElement), this function returns true only if this layerable has its visibility set to true and the parent layerable's realVisibility returns true.
If this layerable doesn't have a direct layerable parent, returns the state of this layerable's visibility.
This function is called by initializeParentPlot, to allow subclasses to react on the setting of a parent plot. This is the case when 0 was passed as parent plot in the constructor, and the parent plot is set at a later time.
For example, QCPLayoutElement/QCPLayout hierarchies may be created independently of any QCustomPlot at first. When they are then added to a layout inside the QCustomPlot, the top level element of the hierarchy gets its parent plot initialized with initializeParentPlot. To propagate the parent plot to all the children of the hierarchy, the top level element then uses this function to pass the parent plot on to its child elements.
Sets the parent plot of this layerable. Use this function once to set the parent plot if you have passed 0 in the constructor. It can not be used to move a layerable from one QCustomPlot to another one.
Note that, unlike when passing a non-null parent plot in the constructor, this function does not make parentPlot the QObject-parent of this layerable. If you want this, call QObject::setParent(parentPlot) in addition to this function.
Further, you will probably want to set a layer (setLayer) after calling this function, to make the layerable appear on the QCustomPlot.
The parent plot change will be propagated to subclasses via a call to parentPlotInitialized so they can react accordingly (e.g. also initialize the parent plot of child layerables, like QCPLayout does).
Sets the parent layerable of this layerable to parentLayerable. Note that parentLayerable does not become the QObject-parent (for memory management) of this layerable.
The parent layerable has influence on the return value of the realVisibility method. Only layerables with a fully visible parent tree will return true for realVisibility, and thus be drawn.
Moves this layerable object to layer. If prepend is true, this object will be prepended to the new layer's list, i.e. it will be drawn below the objects already on the layer. If it is false, the object will be appended.
Returns true on success, i.e. if layer is a valid layer.
Sets the QCPainter::setAntialiasing state on the provided painter, depending on the localAntialiased value as well as the overrides QCustomPlot::setAntialiasedElements and QCustomPlot::setNotAntialiasedElements. Which override enum this function takes into account is controlled via overrideElement.
The documentation for this class was generated from the following files:
Expands this range such that otherRange is contained in the new range. It is assumed that both this range and otherRange are normalized (see normalize).
If otherRange is already inside the current range, this function does nothing.
Returns a sanitized version of the range. Sanitized means for logarithmic scales, that the range won't span the positive and negative sign domain, i.e. contain zero. Further lower will always be numerically smaller (or equal) to upper.
If the original range does span positive and negative sign domains or contains zero, the returned range will try to approximate the original range as good as possible. If the positive interval of the original range is wider than the negative interval, the returned range will only contain the positive interval, with lower bound set to rangeFac or rangeFac *upper, whichever is closer to zero. Same procedure is used if the negative interval is wider than the positive interval, this time by changing the upper bound.
Checks, whether the specified range is within valid bounds, which are defined as QCPRange::maxRange and QCPRange::minRange. A valid range means:
range bounds within -maxRange and maxRange
range size above minRange
range size below maxRange
Field Documentation
const double QCPRange::minRange = 1e-280
static
Minimum range size (upper - lower) the range changing functions will accept. Smaller intervals would cause errors due to the 11-bit exponent of double precision numbers, corresponding to a minimum magnitude of roughly 1e-308.
Maximum values (negative and positive) the range will accept in range-changing functions. Larger absolute values would cause errors due to the 11-bit exponent of double precision numbers, corresponding to a maximum magnitude of roughly 1e308. Since the number of planck-volumes in the entire visible universe is only ~1e183, this should be enough.
