Marginal distributions for heatmap

[hmeine]

heatmap() is very useful for confusion matrices and the like, in particular with annot = True. Still, it would be nice to be able to see the marginal distributions as well, much like with jointplot(). That would allow to assess imbalanced distributions much better.

[daviddavo]

I tried doing it with JointGrid with kind of successful results

The only problem is that the heatmap has kind of an offset with the marginal distributions...

If anyone knows how to fix it, feel welcome to contribute

Option 1: Without setting limits

# _toplot = pd.crosstab(...)

grid = sns.JointGrid()

sns.heatmap(_toplot, ax=grid.ax_joint, cbar=False, annot=True, fmt='g')
sns.barplot(_toplot.sum(), ax=grid.ax_marg_x)
sns.barplot(_toplot.sum(axis=1), ax=grid.ax_marg_y, orient='h')

grid.ax_joint.set_xlabel('Día de votación')
grid.ax_joint.set_ylabel('Día de creación')
grid.ax_joint.xaxis.set_tick_params(rotation=30)
grid.ax_joint.yaxis.set_tick_params(rotation=0)

Option 2: With limits

I tried adding these two lines to "fix" the heatmap, but then part of the countplot gets cut T.T

grid.ax_joint.set_xlim(0, 7)
grid.ax_joint.set_ylim(0, 7)

[daviddavo]

Done! I just had to remove .5 to every x in the bar

# _toplot = pd.crosstab(...)

grid = sns.JointGrid()

sns.heatmap(_toplot, ax=grid.ax_joint, cbar=False, annot=True, fmt='g')
sns.barplot(_toplot.sum(), ax=grid.ax_marg_x, width=1)
sns.barplot(_toplot.sum(axis=1), ax=grid.ax_marg_y, orient='h', width=1)
_off = .5

# Fix x
_xmin, _xmax = grid.ax_joint.get_xlim()
grid.ax_joint.set_xlim(_xmin+_off, _xmax+_off)
for bar in grid.ax_marg_x.containers[0]:
    bar.set_x(bar.get_x() + _off)

# Fix y
_ymin, _ymax = grid.ax_joint.get_ylim()
grid.ax_joint.set_ylim(_ymin+_off, _ymax+_off)
for bar in grid.ax_marg_y.containers[0]:
    bar.set_y(bar.get_y() + _off)

grid.ax_joint.xaxis.set_tick_params(rotation=30)
grid.ax_joint.yaxis.set_tick_params(rotation=0)

grid.ax_joint.set_xlabel('Día de votación')
grid.ax_joint.set_ylabel('Día de creación')