python - Matplotlib 堆叠条形图，具有类似于 pgfplots 的单独边缘颜色

Question

我有一些堆叠的条形图，可能看起来像这样

但我真的很喜欢跟随其余颜色的边缘颜色，例如像 pgfplots

这在 Matplotlib 中可能(在合理范围内)吗？

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

test1 = np.array([51, 13.8, 15.5, np.NaN])
test2 = np.array([40.3, 30.4, 13.8, 15.5])

df = pd.DataFrame(dict(test1 = test1,
                       test2 = test2),
                  columns = ["test1","test2"]).T

ax = df.plot.barh(stacked = True, cmap = "coolwarm", edgecolor = "black", lw = 1, width = 0.8, figsize = (6,4))
plt.show()

Answer 1

首先，条形图的边缘是黑色的，因为您在调用 df.plot.barh 时设置了 edgecolor = "black"。删除它意味着不会有边缘颜色。您需要将每个条形的边缘颜色设置为条形的面颜色

您可以通过迭代矩形补丁(使用 ax.patches 获得)并使用 get_facecolor() 将边缘颜色设置为面部颜色来完成此操作。和 set_edgecolor() .

test1 = np.array([51, 13.8, 15.5, np.NaN])
test2 = np.array([40.3, 30.4, 13.8, 15.5])

df = pd.DataFrame(dict(test1 = test1,
                       test2 = test2),
                  columns = ["test1","test2"]).T

ax = df.plot.barh(stacked = True, cmap = "coolwarm", lw = 1.5, width = 0.8, figsize = (6,4))

for rect in ax.patches:
    facecolor = list(rect.get_facecolor())
    rect.set_edgecolor(facecolor)

    facecolor[-1] = 0.5  # reduce alpha value of facecolor, but not of edgecolor
    rect.set_facecolor(facecolor)

plt.show()

这给出: