Probplot tick and ticklabel formatting under py2.7

[austinorr]

probplot ticks and ticklabels render incorrectly for plottype = 'prob' under py2.7.11

code as follows:

import warnings
warnings.simplefilter('ignore')

import numpy
from matplotlib import pyplot
import seaborn

import probscale
clear_bkgd = {'axes.facecolor':'none', 'figure.facecolor':'none'}
seaborn.set(style='ticks', context='notebook', rc=clear_bkgd)
%matplotlib inline
In [2]:
numpy.random.seed(0)
sample = numpy.random.normal(loc=4, scale=2, size=37)
In [3]:
fig, (ax1, ax2, ax3) = pyplot.subplots(nrows=3, figsize=(5.5, 7))

probscale.probplot(sample, ax=ax1, plottype='pp', xlabel='Percentiles')
probscale.probplot(sample, ax=ax2, plottype='qq', xlabel='Quantiles')
probscale.probplot(sample, ax=ax3, plottype='prob', xlabel='Probabilities')

ax2.set_xlim(left=-2.5, right=2.5)
ax3.set_xlim(left=0.5, right=99.5)
fig.tight_layout()
seaborn.despine(fig=fig)

Out [3]:

Is it possible to manually format tick labels, or is their formatting handled completely internally? I was not able to retrieve them and apply string formatting to them manually.

[phobson]

I'm not inclined to spend any effort supporting legacy python, but retrieving the labels and applying string formatting is not how matplotlib tick labels are formatted. Instead you pass a Formatter class to e.g., ax.xaxis.set_major_formatter. String formatting also won't help rendering issues

The crux of the probscale formatter is here:
https://github.com/phobson/mpl-probscale/blob/master/probscale/formatters.py#L75

And here's an example of using matplotlib's FuncFormatter function to create a custom formatter class:
http://matplotlib.org/examples/pylab_examples/custom_ticker1.html