CODEWITHSUNDEEP
pythonmatplotliblabeldata-visualizationheatmap
AdiT
AdiT

Reputation: 549

How to center ticks and labels in a heatmap

I am plotting a heatmap using matplotlib like the figure below:

matplotlib heatmap

The plot is constructed via the code below:

C_range = 10. ** np.arange(-2, 8)
gamma_range = 10. ** np.arange(-5, 4)

confMat=np.random.rand(10, 9)

heatmap = plt.pcolor(confMat)

for y in range(confMat.shape[0]):
    for x in range(confMat.shape[1]):
        plt.text(x + 0.5, y + 0.5, '%.2f' % confMat[y, x],
                horizontalalignment='center',
                verticalalignment='center',)


plt.grid()
plt.colorbar(heatmap)
plt.subplots_adjust(left=0.15, right=0.99, bottom=0.15, top=0.99)
plt.ylabel('Cost')
plt.xlabel('Gamma')

plt.xticks(np.arange(len(gamma_range)), gamma_range, rotation=45,)
plt.yticks(np.arange(len(C_range)), C_range, rotation=45)
plt.show()

I need to center the ticks and labels on both axes. Any ideas?

Upvotes: 3

Views: 10781

Answers (1)

Andras Deak -- Слава Україні
Andras Deak -- Слава Україні

Reputation: 35094

For your specific code the simplest solution is to shift your tick positions by half a unit separation:

import numpy as np
import matplotlib.pyplot as plt

C_range = 10. ** np.arange(-2, 8)
gamma_range = 10. ** np.arange(-5, 4)

confMat=np.random.rand(10, 9)

heatmap = plt.pcolor(confMat)

for y in range(confMat.shape[0]):
    for x in range(confMat.shape[1]):
        plt.text(x + 0.5, y + 0.5, '%.2f' % confMat[y, x],
                horizontalalignment='center',
                verticalalignment='center',)


#plt.grid() #this will look bad now
plt.colorbar(heatmap)
plt.subplots_adjust(left=0.15, right=0.99, bottom=0.15, top=0.99)
plt.ylabel('Cost')
plt.xlabel('Gamma')

plt.xticks(np.arange(len(gamma_range))+0.5, gamma_range, rotation=45,)
plt.yticks(np.arange(len(C_range))+0.5, C_range, rotation=45)
plt.show()

result

As you can see, in this case you need to turn off the grid, otherwise it will overlap with your squares and clutter up your plot.

Upvotes: 5

Related Questions