部分相关可视化

你的意思是这样的:

import matplotlib.pyplot as plt
partial_corr_array = np.array([[ 1., 0.52, 0.15, 0.91], [ 0.52, 1., 0.25, -0.25], [ 0.15,  0.25,  1.  , -0.06], [ 0.91, -0.25, -0.06,  1.  ]])
plt.matshow(partial_corr_array)
plt.colorbar()

输出

图2 基于 this

import seaborn as sns
f, ax = plt.subplots(figsize=(6, 5))
partial_corr_array = np.array([[ 1., 0.52, 0.15, 0.91], [ 0.52, 1., 0.25, -0.25], [ 0.15,  0.25,  1.  , -0.06], [ 0.91, -0.25, -0.06,  1.  ]])
mask = np.zeros_like(partial_corr_array, dtype=np.bool)
mask[np.triu_indices_from(mask)] = True
cmap = sns.diverging_palette(30, 10, as_cmap=True)
sns.heatmap(partial_corr_array, mask=mask, cmap=cmap, vmax=1, center=0,
            square=True, linewidths=1, cbar_kws={"shrink": 0.9}, annot=True)

输出

seaborn.heatmap 提供的正是这个,使用 fmt 控制面值字符串格式,并使用 annot 要控制是否显示面值(默认为False):

DataFrame.corr

In [28]: import seaborn as sns
In [39]: sns.heatmap(df.corr(), fmt='0.2f', annot=True)
Out[39]: <matplotlib.axes._subplots.AxesSubplot at 0xd961048>