Drawing a colorbar with pg.mplviewer.drawModel

[Coastal0]

Hi guys,

I'm trying to draw a colorbar with the drawModel command. To my mind, the interpolation tutorial suggests it's possible (by means of showCbar = False) however setting this to 'True' does nothing.

I introduced a colorbar through the matplotlib which seems to work just fine.

dataplot = pg.mplviewer.drawModel(ax = ax, mesh = meshInv, data = datInv, colorBar=True, cmap='jet_r', showMesh=0, cMin = 10, cMax = 2000)
cb = plt.colorbar(dataplot)

Should drawModel be taking the showCbar command?

[florian-wagner]

No, drawModel only draws the model into a specified axes. Please use pg.viewer.showMesh or pg.show for high-level plotting, which will take the argument colorBar=True.

import pygimli as pg

mesh = pg.createGrid([0,1,2], [0,1,2])
data = pg.randn(mesh.cellCount())

pg.show(mesh, data, colorBar=True)
pg.show(mesh, data, colorBar=True, orientation="vertical")

[florian-wagner]

If you are preparing more involved figures, you might wanna have a look at:

• https://www.pygimli.org/pygimliapi/_generated/pygimli.mplviewer.html?highlight=colorbar#pygimli.mplviewer.createColorBarOnly
• https://www.pygimli.org/pygimliapi/_generated/pygimli.mplviewer.html?highlight=colorbar#pygimli.mplviewer.findColorBar
• https://www.pygimli.org/pygimliapi/_generated/pygimli.mplviewer.html?highlight=colorbar#pygimli.mplviewer.createColorBar

[Coastal0]

On a related (but not new topic-worthy) note;

How on earth are you supposed to access these ticklabels?
I simply want them in a different format (i.e. not 10^2,3 etc), yet I can't access them in the same way I would if it were simply matplotlib.

This code below, for example, (and based on my earlier question), does exactly what I'd like it to do;

dataplot = pg.mplviewer.drawModel(...parameters...)
cb = plt.colorbar(dataplot, orientation = 'horizontal', format = '%.i',aspect = 50)

But when creating the colorbar through pg.show, I can't see how to access the same format options.

ax, cb = pg.show(meshERT, resBulk, colorBar=True, cmap='jet_r', showMesh=0)

I can turn them to integers with the following;

cb.ax.xaxis.set_ticklabels(cb.get_ticks().astype(int), minor = False)

However this shows them all, when I just want to change the existing set of shown numbers.

Reading through Florian's links to colorbars, I can't see a way of changing the format. I'm at the point now where it's better to ask than keep getting nowhere alone...!

[halbmy]

The new 10^x colorbars (from mpl 2.1 or 2.2 on) are indeed a bit nasty.

It is true, pg.mplviewer.createColorbar (https://www.pygimli.org/pygimliapi/_generated/pygimli.mplviewer.html?highlight=colorbar#pygimli.mplviewer.createColorBar) does not (unlike createColorBarOnly) pass additional keyword arguments to mpl.colorbar (maybe it should), but it returns the colorbar object that can be manipulated afterwards.

[Coastal0]

What actor (If that's the right term?) does pg.show put out?

For example, in the earlier code, the plt.colorbar command takes the pg.drawModel output, however I can't seem to find the similar object from pg.show

I guess it's a patch-style object, but I'm just not sure how to get at it.

[halbmy]

pg.show alias pg.viewer.show redirects to pg.viewer.showMesh and returns the axis and the colorbar object: ax, cbar = pg.show(mesh, property, colorBar=True)