manipulate tick labels on a colorbar

Question

I want to access the tick labels on my matplotlib colobar, so that I can manipulate them.

My starting labels may be [-2,-1,0,1,2] for example.

I have used:

locs,oldlabels = plt.xticks()
newlabels = ['a','b','c','d','e']
plt.xticks(locs, newlabels)

This works. But I don't want to manually write in the new labels. I want to access the oldlabels, so that I can have the newlabels as say [2*(-2), 2*(-1), 2*0, 2*1, 2*2].

I just don't know how to 'get at' the oldlabels. I googled everything and tried lots of things, but I'm doing something fundamentally wrong.

I tried to print oldlabels[0], but I get Text(0,0,u'\u22122.0').

---

EDIT:

I'm currently doing:

new_labels = [1,2,3,4,5,6,7,8,9]
colorbarname.ax.set_xticklabels(new_labels)

which works. But I want to set them as 2 x their old value. How can I do this automatically? I need to extract the old label values, multiply by (say) 2, update the axis labels with the new values.

Answer 1

If your data is not confined to [0,1], I'd recommend using a norm when you pass the data to the colormap instead of changing the data and relabeling the colorbar: http://matplotlib.org/api/cm_api.html?highlight=norm%20colormap#matplotlib.cm.ScalarMappable.norm

However, you can relabel the colorbar by manipulating the underlying axis directly:

import numpy as np
import pylab as plt

A = np.random.random((10,10))

plt.subplot(121)
plt.imshow(A,interpolation='nearest')

cb = plt.colorbar()
oldlabels = cb.ax.get_yticklabels()
print(map(lambda x: x.get_text(),oldlabels))
newlabels = map(lambda x: str(2 * float(x.get_text())), oldlabels)
print(newlabels)
cb.ax.set_yticklabels(newlabels)
plt.show()

oh, and now I find the matplotlib gallery example, nearly the same: http://matplotlib.org/examples/pylab_examples/colorbar_tick_labelling_demo.html