python - 如何将离散值映射到 seaborn 中的热图？

Question

我正在尝试使用 seaborn 在热图中绘制离散值。这是我要绘制的列表:

xa = [[5, 4, 4, 4, 13, 4, 4],
 [1, 9, 4, 3, 9, 1, 4],
 [4, 1, 7, 1, 5, 3, 7],
 [1, 9, 4, 3, 9, 5, 4],
 [2, 1, 4, 1, 9, 4, 3],
 [9, 4, 8, 1, 7, 1, 9],
 [4, 8, 1, 7, 1, 4, 8]]

这是我用来绘制热图的代码:

import numpy as np
import seaborn as sns
from matplotlib.colors import ListedColormap
data = np.asarray(xa)
sns.heatmap( data,cmap=ListedColormap(['green', 'yellow', 'red']))

我的问题是如何将每个数字绘制成特定颜色。值的范围将从 1 到 17。所以 17 种不同的颜色，每个数字一种。我确实阅读了其他一些答案，但都没有谈到如何为数字分配特定值。谢谢!

Answer 1

如果我没理解错的话，你可以这样做:

import numpy as np
from matplotlib import pyplot as plt
import matplotlib.colors as c
data = np.asarray(xa)
colors = {"white":1, "gray":2, "yellow":3, "lightgreen":4, "green":5, "lightblue":6, "blue":7, "lightcoral":8, "red":9, "brown":10,
          "violet":11, "blueviolet":12, "indigo":13, "khaki":14, "orange":15, "pink":16, "black":17}
l_colors = sorted(colors, key=colors.get)
cMap = c.ListedColormap(l_colors)
fig, ax = plt.subplots()
ax.pcolor(data[::-1], cmap=cMap, vmin=1, vmax=len(colors))
# plt.axis('off') # if you don't want the axis
plt.show()

每个数字对应一种颜色，从 1(白色)、2(灰色)到 17(黑色)。如您所见，图像中没有黑色，因为数组中没有 17，并且颜色图未标准化。

或者使用seaborn:

data = np.asarray(xa)
colors = {"white":1,"gray":2,"yellow":3,"lightgreen":4, "green":5, "lightblue":6, "blue":7, "lightcoral":8, "red":9, "brown":10,
          "violet":11, "blueviolet":12,"indigo":13, "khaki":14, "orange":15, "pink":16, "black":17}
l_colors = sorted(colors, key=colors.get)
cMap = c.ListedColormap(l_colors)
sns.heatmap(data,cmap=l_colors, vmin=1, vmax=len(colors))

如果您想要图例上的所有刻度，请添加:

ax = sns.heatmap(data,cmap=l_colors, vmin=1, vmax=len(colors))
colorbar = ax.collections[0].colorbar
colorbar.set_ticks([1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17])