Recent version of matplotlib won't plot with colorbar

[cgadal]

As mentioned in issue #1006 and #1009, trying to view a CellVariable using

viewer = MatplotlibViewer(vars=varname, datamin=-0, datamax=0, title="Grid2D test", axes=ax)

or

viewer = Viewer(vars=phi, datamin=-1, datamax=1.)

will result in

ValueError: Unable to determine Axes to steal space for Colorbar. Either provide the *cax* argument to use as the Axes for the Colorbar, provide the *ax* argument to steal space from it, or add *mappable* to an Axes.

See for example https://pages.nist.gov/fipy/en/latest/generated/examples.diffusion.circle.html#module-examples.diffusion.circle.

This is due to a change in recent versions of matplotlib where fig.colorbar will not guess anymore from which ax it will steal space. See for example matplotlib/matplotlib#23973.

Temporary fixes:

• either roll back to matplotlib 5.1
• use viewer without the colorbar as viewer = MatplotlibViewer(vars=phi, datamin=-0, datamax=0, title="Grid2D test", axes=ax, colorbar=None)

change in fipy code:

I am new to fipy, but I suspect that passing explicitly the ax in this part of the code would solve the problem:

fipy/fipy/viewers/matplotlibViewer/abstractMatplotlibViewer.py

Lines 120 to 123 in e7afd1a

	if colorbar:
	self._colorbar = self.fig.colorbar(mappable=self._mappable,
	orientation=colorbar,
	label=self.vars[0].name)

maybe as:

 if colorbar: 
     self._colorbar = self.fig.colorbar(mappable=self._mappable, 
                                        orientation=colorbar, 
                                        label=self.vars[0].name, 
                                        ax=self.axes) 

I have not tried yet tough.
Tried it and it works perfectly it seems. Will open a pull request.