0

I am generating a heatmap using seaborn which has a logarithmic scale. How can I change the colorbar labels from scientific notation to plain number.

import math
from matplotlib.colors import LogNorm
vmax=2
vmin=0.5
center = (vmax+vmin)/2
log_norm = LogNorm(vmin=vmin, vmax=vmin)
cbar_ticks = [0.5, 0.66, 1, 1.5, 2]
ax = sns.heatmap(corr, square=True, mask=mask, cmap=cmap_type, linewidths=.5, vmax=vmax, vmin=vmin, norm=log_norm, cbar_kws={"ticks": cbar_ticks}, center=center)

generated graph

Edit:

The following code fixes the problem with scientific notation:

    import matplotlib.ticker as tkr
    formatter = tkr.ScalarFormatter(useMathText=True)
    formatter.set_scientific(False)
    cbar_kws={"ticks": cbar_ticks, "format": formatter}

but norm=log_norm is adding additional ticks (Please see image below). How do I remove these?

Updated graph

wolfe1947
  • 25
  • 5

2 Answers2

0

seaborn plots are wrappers around matplotlib so I would suggest this

ax.ticklabel_format(style='plain')
JSharm
  • 1,117
  • 12
  • 11
0

The following codes helped.

For the first part of the problem adding "format" in cbar_kws parameters helped. (https://stackoverflow.com/a/35419495/8881141):

import matplotlib.ticker as tkr
formatter = tkr.ScalarFormatter(useMathText=True)
formatter.set_scientific(False)

ax = sns.heatmap(corr, square=True, mask=mask, cmap=cmap_type, linewidths=.5, vmax=vmax, vmin=vmin, norm=log_norm, cbar_kws={"ticks": cbar_ticks, "format": formatter}, center=center)

For the second part, which is caused due to a bug in matplotlib, adding the following removes the minor ticks. (https://stackoverflow.com/a/65676007/11897903):

ax.collections[0].colorbar.ax.yaxis.set_ticks([], minor=True)
wolfe1947
  • 25
  • 5