两个混淆矩阵图有不同的大小

Question

我试图在同一图像中绘制混淆矩阵，但它们的大小不同。

代码如下：

fig, ax = plt.subplots(nrows=1, ncols=2, figsize = (18,8))
fig.suptitle('Matriz de Confusão')


skplt.metrics.plot_confusion_matrix(y_test, y_pred_log, normalize=True, ax=ax[0], title=('Regressão Logística'))
skplt.metrics.plot_confusion_matrix(y_test, y_pred_tree, normalize=True, ax=ax[1], title=('Árvore de decisão'))
ax[0].xaxis.set_ticklabels(['Normal', 'Fraude']); ax[0].yaxis.set_ticklabels(['Normal', 'Fraude']);
ax[1].xaxis.set_ticklabels(['Normal', 'Fraude']); ax[1].yaxis.set_ticklabels(['Normal', 'Fraude']);


plt.show()

这就是我得到的：

如何更改第二个绘图的大小？
另外要是能把多余的彩条删掉就好了

Answer 1

您应该定义需要放置颜色条的轴。您可以查看作为参考。
将这些概念应用到您的案例中会产生类似这样的结果：

import matplotlib.pyplot as plt
import numpy as np


M1 = np.random.rand(2, 2)
M2 = np.random.rand(2, 2)


fig, ax = plt.subplots(1, 2, figsize = (18, 8))
plt.subplots_adjust(right = 0.77)
cbar_ax_1 = fig.add_axes([0.8, 0.1, 0.04, 0.8])
cbar_ax_2 = fig.add_axes([0.9, 0.1, 0.04, 0.8])

im_1 = ax[0].imshow(M1, cmap = 'magma')
im_2 = ax[1].imshow(M2, cmap = 'magma')

plt.colorbar(im_1, cax = cbar_ax_1)
plt.colorbar(im_2, cax = cbar_ax_2)

plt.show()

如果您只想要一个颜色条，明智的做法是根据两个矩阵的值对唯一颜色条进行归一化：

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import Normalize
from matplotlib import cm


M1 = np.random.rand(2, 2)
M2 = np.random.rand(2, 2)


fig, ax = plt.subplots(1, 2, figsize = (18, 8))
plt.subplots_adjust(right = 0.87)
cbar_ax = fig.add_axes([0.9, 0.1, 0.04, 0.8])
norm = Normalize(vmin = min(np.min(M1), np.min(M2)), vmax = max(np.max(M1), np.max(M2)))
cmap = cm.magma

im_1 = ax[0].imshow(M1, cmap = cmap)
im_2 = ax[1].imshow(M2, cmap = cmap)

plt.colorbar(cm.ScalarMappable(norm = norm, cmap = cmap), cax = cbar_ax)

plt.show()

两个混淆矩阵图有不同的大小

Two confusion matrix plots have different sizes

python

plot

matplotlib

confusion-matrix

scikit-learn