Heatmap only supports equal row and column

Description

heatmap only works for equal row and column.

Steps to reproduce

import proplot as plot
import pandas as pd

data = pd.DataFrame(index=list('abcdefghij'), columns=list('lmn'))
data = data.fillna(0)
print(data.values.shape)

# Covariance matrix plot
f, ax = plot.subplots(axwidth=4.5)
m = ax.heatmap(
    data, cmap='ColdHot', vmin=-1, vmax=1, N=100,
    lw=0.5, edgecolor='k', labels=True, labels_kw={'weight': 'bold'},
    clip_on=False, # turn off clipping so box edges are not cut in half
)
ax.format(
    suptitle='Heatmap demo', title='Table of correlation coefficients', alpha=0, linewidth=0,
    xloc='top', yloc='right', yreverse=True, ticklabelweight='bold',
    ytickmajorpad=4,  # the ytick.major.pad rc setting; adds extra space
)

Expected behavior (generated by seaborn):

import seaborn as sns
import pandas as pd

data = pd.DataFrame(index=list('abcdefghij'), columns=list('lmn'))
data = data.fillna(0)
sns.set()
ax = sns.heatmap(data, cmap="YlGn")

Actual behavior:

  File "E:\miniconda3\envs\satpy\lib\site-packages\proplot\subplots.py", line 217, in _iterator
    ret.append(func(*args, **kwargs))
  File "E:\miniconda3\envs\satpy\lib\site-packages\proplot\axes.py", line 1275, in heatmap
    obj = self.pcolormesh(*args, **kwargs)
  File "E:\miniconda3\envs\satpy\lib\site-packages\proplot\wrappers.py", line 3083, in _wrapper
    return driver(self, func, *args, **kwargs)
  File "E:\miniconda3\envs\satpy\lib\site-packages\proplot\wrappers.py", line 551, in standardize_2d
    f'Input shapes x {x.shape} and y {y.shape} must match '
ValueError: Input shapes x (10,) and y (3,) must match Z centers (10, 3) or Z borders (11, 4).

Proplot version

0.5.0

[bradyrx]

I think ax.heatmap just doesn’t take raw pandas dataframes. If you do data.values it works.

Here’s an example with some random data inside and calling data.columns and data.index for proper labeling:

import proplot as plot
import pandas as pd


data = pd.DataFrame(index=list('abcdefghij'), columns=list('lmn'))
data[:] = np.random.rand(10, 3)

f, ax = plot.subplots(axwidth=4.5)
m = ax.heatmap(data.columns, data.index,
    data, cmap='ColdHot', vmin=-1, vmax=1, N=100,
    lw=0.5, edgecolor='k', labels=True, labels_kw={'weight': 'bold'},
    clip_on=False, # turn off clipping so box edges are not cut in half
)
ax.format(
    suptitle='Heatmap demo', title='Table of correlation coefficients', alpha=0, linewidth=0,
    xloc='top', yloc='right', yreverse=True, ticklabelweight='bold',
    ytickmajorpad=4,  # the ytick.major.pad rc setting; adds extra space
)

[lukelbd]

This is fixed by #159. Note you can also make the grid boxes square by passing aspect='equal' to format() (which calls Axes.set_aspect()) although #159 also makes “equal” aspect ratio the default behavior for heatmaps, just like it is the default for imshow plots.