source: orange/docs/extend-widgets/rst/plots.rst @ 11408:c2d2400b6a90

Revision 11408:c2d2400b6a90, 6.8 KB checked in by Ales Erjavec <ales.erjavec@…>, 13 months ago (diff)

Fixes for Widgets Development documentation.

RevLine 
[11049]1###########################
2Graphing and Orange Widgets
3###########################
4 
5The most fun widgets are of course those that include graphics. For
6this we can use the classes for
7drawing data plots as provided in Qwt library and PyQwt
8interface. This method is described in :doc:`Graphing <graphing>`.
9
10However, Orange provides a new plotting interface via the :obj:`OrangeWidgets.plot`
11module. The OWPlot class use Qt's graphics framework and was written specifically for
12Orange, so it contains methods more suitable to its data structures. It also provides
13most methods also found in the Qwt-based OWGraph, so switching from one base
14to another is quite easy.
15
16Creating a new plot using OWPlot is described in :doc:`Plots with OWPlot <plotsbasic>`.
17This topic tries to show the similarities between the plot module and the old Qwt-based graph.
18
19*****
20Plots
21*****
22
23Let us construct a widget with a following appearance:
24
25.. image:: learningcurve-owplot.png
26
27There are two new elements from our previous incarnation of
28a learning curve widget: a control with a list of classifiers, and a
29graph with a plot of learning curves. From a list of classifiers we
30can select those to be displayed in the plot.
31
32The widget still provides learning curve table, but this is now
33offered in a tabbed pane together with a graph. The code for
34definition of the tabbed pane, and initialization of the graph is::
35
36    # start of content (right) area
37    tabs = OWGUI.tabWidget(self.mainArea)
38   
39    # graph widget
40    tab = OWGUI.createTabPage(tabs, "Graph")
41    self.graph = OWPlot(tab)
42    self.graph.setAxisAutoScale(QwtPlot.xBottom)
43    self.graph.setAxisAutoScale(QwtPlot.yLeft)
44    tab.layout().addWidget(self.graph)
45    self.setGraphGrid()
46
47:obj:`OWPlot` is a convenience subclass of QGraphicsView and is imported from plot.owplot module.
48For the graph, we use :obj:`set_axis_autoscale` to
49request that the axis are automatically set in regard to the data that
50is plotted in the graph. We plot the graph in using the following
51code::
52
53    def drawLearningCurve(self, learner):
54        if not self.data: return
55        curve = self.graph.add_curve(learner.name, xData=self.curvePoints, yData=learner.score, autoScale=True)
56
57        learner.curve = curve
58        self.setGraphStyle(learner)
59        self.graph.replot()
60
61This is simple. We store the curve returned from :obj:`add_curve` with a
62learner, and use a trick allowed in Orange that we can simply store
63this as a new attribute to the learning object. In this way, each learner
64also stores the current scores, which is a list of numbers to be
65plotted in the graph. The details on how the plot is set are dealt
66with in :obj:`setGraphStyle` function::
67
68    def setGraphStyle(self, learner):
69        curve = learner.curve
70        if self.graphDrawLines:
71            curve.set_style(OWCurve.LinesPoints)
72        else:
73            curve.set_style(OWCurve.Points)
74        curve.set_symbol(OWPoint.Ellipse)
75        curve.set_point_size(self.graphPointSize)
76        curve.set_color(self.graph.color(OWPalette.Data))
77        curve.set_pen(QPen(learner.color, 5))
78
79Notice that the color of the plot line that is specific to the
80learner is stored in its attribute :obj:`color`
81(:obj:`learner.color`). Who sets it and how? This we discuss in
82the following subsection.
83
[11408]84************************
85Colors in Orange Widgets
86************************
[11049]87
88Uniform assignment of colors across different widget is an
89important issue. When we plot the same data in different widgets, we
90expect that the color we used in a consistent way; for instance data
91instances of one class should be plotted in scatter plot and parallel
92axis plot using the same color. Developers are thus advised to use
93:obj:`ColorPaletteHSV`, which is provided as a method within
94:obj:`OWWidget` class. :obj:`ColorPaletteHSV` takes an
95integer as an attribute, and returns a list of corresponding number of
96colors. In our learning curve widget, we use it within a function that
97sets the list box with learners::
98
99    def updatellb(self):
100        self.blockSelectionChanges = 1
101        self.llb.clear()
102        colors = ColorPaletteHSV(len(self.learners))
103        for (i,lt) in enumerate(self.learners):
104            l = lt[1]
105            item = QListWidgetItem(ColorPixmap(colors[i]), l.name)
106            self.llb.addItem(item)
107            item.setSelected(l.isSelected)
108            l.color = colors[i]
109        self.blockSelectionChanges = 0
110
111The code above sets the items of the list box, where each item
112includes a learner and a small box in learner's color, which is in
113this widget also used as a sort of a legend for the graph. This box is
114returned by :obj:`ColorPixmap` function defined in
115:obj:`OWColorPalette.py`. Else, the classifier's list box control is
116defined in the initialization of the widget using::
117
118    self.cbox = OWGUI.widgetBox(self.controlArea, "Learners")
119    self.llb = OWGUI.listBox(self.cbox, self, "selectedLearners", selectionMode=QListWidget.MultiSelection, callback=self.learnerSelectionChanged)
120
121    self.llb.setMinimumHeight(50)
122    self.blockSelectionChanges = 0
123
124Now, what is this :obj:`blockSelectionChanges`? Any time
125user makes a selection change in list box of classifiers, we want to
126invoke the procedure called
127:obj:`learnerSelectionChanged`. But we want to perform
128actions there when changes in the list box are invoked from clicking
129by a user, and not by changing list box items from a program. This is
130why, every time we want :obj:`learnerSelectionChanged` not to
131perform its function, we set :obj:`self.blockSelectionChanges`
132to 1.
133
134In our widget, :obj:`learnerSelectionChanged` figures out
135if any curve should be removed from the graph (the user has just
136deselected the corresponding item in the list box) or added to the
137graph (the user just selected a learner)::
138
139    def learnerSelectionChanged(self):
140        if self.blockSelectionChanges: return
141        for (i,lt) in enumerate(self.learners):
142            l = lt[1]
143            if l.isSelected != (i in self.selectedLearners):
144                if l.isSelected: # learner was deselected
145                    l.curve.detach()
146                else: # learner was selected
147                    self.drawLearningCurve(l)
148                self.graph.replot()
149            l.isSelected = i in self.selectedLearners
150
[11408]151The complete code of this widget is available :download:`here <OWLearningCurve_plot.py>`.
[11049]152This is almost like a typical
153widget that is include in a standard Orange distribution, with a
154typical size just under 300 lines. Just some final cosmetics is needed
155to make this widget a standard one, including setting some graph
156properties (like line and point sizes, grid line control, etc.) and
157saving the graph to an output file.
158
159***************
160Further reading
161***************
162
163This tutorial only shows one kind of a plot.
164If you wish to use different and more complicated plots,
165refer to the :doc:`Plot module documentation <OrangeWidgets.plot>`.
Note: See TracBrowser for help on using the repository browser.