Heatmap: Specifying number of ticks or tick locations in cmap

First off, thanks so much @mwaskom for this great library. We use it constantly at my work place!

Q. I could not find an option to change the ticks locations (or the number of ticks) for the colormap in a heatmap. That is, I’m not talking about the x or y axis, but the ticks in the cmap itself.

Any suggestions on how to do it?

E.g. for:

# Draw the heatmap with the mask and correct aspect ratio
g = sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1,
                square=True, linewidths=.5, cbar_kws={"shrink": .9}, ax=ax)

[mwaskom]

You can use the ticks colorbar keyword argument (requires version 0.7.1+):

x = np.random.rand(10, 10)
sns.heatmap(x, cbar_kws=dict(ticks=[.2, .5, .6]))

[nightm4re94]

I have the same issue as @andt88 : With a logarithmic norm, the “ticks” keyword gets ignored entirely. This is the code that produces my heatmap:

log_norm = LogNorm(vmin=1, vmax=1e7)
hm = sb.heatmap(data=matrix_df, xticklabels=10, yticklabels=10, cmap=heatmap_colors, norm=log_norm, cbar=True, cbar_kws={"label": lbl_cycle_count, "ticks": None, "shrink": 1}, vmin=1, vmax=1e7)
hm.set_yticklabels(hm.get_yticklabels(), rotation=0)
hm.set_xticklabels(hm.get_xticklabels(), rotation=90)

But in the result, the ticks of the color bar are shown logarithmically.