提问人：Bas R 提问时间：3/23/2023 更新时间：3/23/2023 访问量：99
使用值数组在 matplotlib pcolor grid 中使用网格线绘制边框
Drawing borders using gridlines in matplotlib pcolor grid by using an array of values
问：
我有两个数组，我使用颜色将它们绘制在单独的节点网格上。一个代表一些聚类，另一个代表其他一些值，我将称之为我的特征。示例代码：
import numpy as np
import matplotlib.pyplot as plt
clusters = np.array([[0,2,1], [0,3,1], [3,3,1]]) # make cluster data
features= np.array([[0,0.4,0.7], [0.1,0.3,0.7], [0.5,0.4,0.8]]) # make data of features
# plot clusters
plt.figure()
plt.pcolor(clusters, cmap='jet') # color in nodes
plt.colorbar(ticks=[i for i in range(0, np.amax(clusters)+1)]) # make colorbar legend per cluster
plt.show()
# plot feature grid
plt.figure()
plt.pcolor(features, cmap='bone', vmin=0, vmax=1) # color in nodes
plt.colorbar() # make colorbar legend
plt.show()
在第二个带有特征数据的灰色网格中，我想通过网格线显示聚类之间的边界。预期结果如下所示，其中红线表示聚类之间的边界：
有没有办法使用簇阵列的数据自动绘制这些网格线？任何帮助将不胜感激！
答：
1赞
Bas R
3/23/2023
#1
这有效：
vlines, hlines = [], []
# loop over array with clusters to obtain positions of vertical and horizontal lines
for row_idx, row in enumerate(clusters): # loop over array rows
for col_idx, col in enumerate(row): # loop over array columns per row
if col_idx+1 < clusters.shape[1]: # skip final column, has no right-side neighbouring node
# save tuple if it requires a vertical line indicating a different cluster right of the node
if clusters[row_idx, col_idx] != clusters[row_idx, col_idx+1]: vlines.append((row_idx, col_idx+1))
if row_idx+1 < clusters.shape[0]: # skip final row, has no bottom neighbouring node
# save a tuple if it requires a horizontal line indicating a different cluster below the node
if clusters[row_idx, col_idx] != clusters[row_idx+1, col_idx]: hlines.append((row_idx+1, col_idx))
# make features plot
plt.figure()
plt.pcolor(features, cmap='bone', vmin=0, vmax=1) # color in nodes
plt.colorbar() # make colorbar legend
for vline in vlines: # plot all vertical lines
plt.axvline(vline[1], ymin=vline[0]/clusters.shape[0], ymax=vline[0]/clusters.shape[0] + 1/clusters.shape[0], color='red')
for hline in hlines: # plot all horizontal lines
plt.axhline(hline[0], xmin=hline[1]/clusters.shape[0], xmax=hline[1]/clusters.shape[0] + 1/clusters.shape[0], color='red')
plt.show()
评论