在seaborn中自定义颜色条 - 热图

Question

I have a heatmap to represent discrete values.

import seaborn as sns
import pandas as pd

data = np.array([[2, 1, 1, 1, 2, 3, 3, 3, 3, 3],
                 [3, 1, 1, 1, 1, 1, 1, 1, 1, 1],
                 [1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
                 [1, 1, 1, 1, 1, 1, 1, 1, 1, 1],])

x_labels = ['a', 'b', 'c', 'd']

dataFrame = pd.DataFrame(data.T)

#get discrete colormap
cmap = colors.ListedColormap(['blue','red','black','yellow'])

ax = sns.heatmap(dataFrame, cmap=cmap, linewidths=.5, linecolor='lightgray')
ax.set_xticklabels(x_labels)
ax.set_yticklabels(['1', '2', '3', '4', '5', '6', '7', '8', '9', '10'])

# Manually specify colorbar labelling after it's been generated
colorbar = ax.collections[0].colorbar
colorbar.set_ticks([1.3, 1.7, 2.2, 2.7])
colorbar.set_ticklabels(['A', 'B', 'C', 'NA'])

plt.show()

How can I specifiy those colors in a way to represent colors = {1: 'A', 2: 'B', 3: 'C', 4: 'NA'} in data

Answer 1

It follows the order you give. Since your code has

cmap = ListedColormap(['red','black','yellow', 'blue'])
colorbar.set_ticklabels(['A', 'B', 'C', 'NA'])

The colors are mapped to these categories. Change the order to, say, ['red','black', 'blue','yellow'] and the labels will change accordingly.

Answer 2

By default, the colormaps are scaled to the range of your data. In your case, your data is [1-3], so matplotlib is scaling your four-color colormap to that range.

Since you, in fact, want to scale the colormap to the range [1-4], you'll have to use the arguments vmin= and vmax= to tell the system what range of data to use.

ax = sns.heatmap(dataFrame, cmap=cmap, linewidths=.5, linecolor='lightgray', vmin=1, vmax=4)