Getting PySide to work with matplotlib

Question

I have tried running the example code on the SciPy website, but I get this error:

Traceback (most recent call last):
  File ".\matplotlibPySide.py", line 24, in <module>
    win.setCentralWidget(canvas)
TypeError: 'PySide.QtGui.QMainWindow.setCentralWidget' called with wrong argument types:
  PySide.QtGui.QMainWindow.setCentralWidget(FigureCanvasQTAgg)
Supported signatures:
  PySide.QtGui.QMainWindow.setCentralWidget(PySide.QtGui.QWidget)

I am building a simple scientific data logger that will eventually be used in commercial applications, so I really need both the LGPL from PySide and plotting functionality. Does anyone have experience on how to get this to work or alternative plotting packages or ideas?

Thanks in advance.

Answer 1

I think you may have posted this on the matplotlib mailing list. But just in case someone else is looking for the answer. The best option is to use the master branch on Github, but if you can't or don't know how to work the Github version you can use the following code to render a plot in PySide.

import numpy as np
from matplotlib import use
use('AGG')
from matplotlib.transforms import Bbox
from matplotlib.path import Path
from matplotlib.patches import Rectangle
from matplotlib.pylab import *
from PySide import QtCore,QtGui

rect = Rectangle((-1, -1), 2, 2, facecolor="#aaaaaa")
gca().add_patch(rect)
bbox = Bbox.from_bounds(-1, -1, 2, 2)

for i in range(12):
    vertices = (np.random.random((4, 2)) - 0.5) * 6.0
    vertices = np.ma.masked_array(vertices, [[False, False], [True, True], [False, False], [False, False]])
    path = Path(vertices)
    if path.intersects_bbox(bbox):
        color = 'r'
    else:
        color = 'b'
    plot(vertices[:,0], vertices[:,1], color=color)

app = QtGui.QApplication(sys.argv)
gcf().canvas.draw()

stringBuffer = gcf().canvas.buffer_rgba(0,0)
l, b, w, h = gcf().bbox.bounds

qImage = QtGui.QImage(stringBuffer, 
                      w,
                      h,
                      QtGui.QImage.Format_ARGB32)

scene = QtGui.QGraphicsScene()
view = QtGui.QGraphicsView(scene)
pixmap = QtGui.QPixmap.fromImage(qImage)
pixmapItem = QtGui.QGraphicsPixmapItem(pixmap)
scene.addItem(pixmapItem)
view.show()

app.exec_()

Answer 2

I had similar goals (LGPL, potential commercial use) and here's how I ended up getting it to work.

Create a matplotlib widget (see here for a more detailed one for PyQt):

import matplotlib

matplotlib.use('Qt4Agg')
matplotlib.rcParams['backend.qt4']='PySide'

from matplotlib.figure import Figure
from matplotlib.backends.backend_qt4agg import FigureCanvasQTAgg as FigureCanvas

class MatplotlibWidget(FigureCanvas):

    def __init__(self, parent=None,xlabel='x',ylabel='y',title='Title'):
        super(MatplotlibWidget, self).__init__(Figure())

        self.setParent(parent)
        self.figure = Figure()
        self.canvas = FigureCanvas(self.figure)
        self.axes = self.figure.add_subplot(111)

        self.axes.set_xlabel(xlabel)
        self.axes.set_ylabel(ylabel)
        self.axes.set_title(title)

In Qt Designer I created a blank widget to hold my plot and then when I __init__ the main window I call setupPlot:

def  setupPlot(self):
    # create a matplotlib widget
    self.DataPlot = MatplotlibWidget()
    # create a layout inside the blank widget and add the matplotlib widget        
    layout = QtGui.QVBoxLayout(self.ui.widget_PlotArea)        
    layout.addWidget(self.DataPlot,1)

Then I call plotDataPoints as needed:

def plotDataPoints(self,x,y):        
    self.DataPlot.axes.clear()
    self.DataPlot.axes.plot(x,y,'bo-')
    self.DataPlot.draw()

Note: this clears and redraws the entire plot every time (since the shape of my data keeps changing) and so isn't fast.

Answer 3

The example that you mention:

http://www.scipy.org/Cookbook/Matplotlib/PySide

works, but you might need to suggest the use of PySide:

...
matplotlib.use('Qt4Agg')
matplotlib.rcParams['backend.qt4']='PySide'
import pylab
...