CODEWITHSUNDEEP
pythonmatplotlibplot
Bogdan Daniel
Bogdan Daniel

Reputation: 2749

Display correlation matrix using axes

I used to plot my correlation matrix like:

corr_matrix = np.corrcoef(W)
    plt.matshow(corr_matrix)
    labels = ['P1', 'P2', 'P3', 'P4', 'P5', 'P6', 'P7', 'P8', 'P9', 'P10', 'P11', 'P12', 'P13', 'P14', 'P15', ]
    plt.xticks(range(0,15),labels)
    plt.yticks(range(0,15),labels)
    plt.colorbar()
    plt.show()

Now I'm working with some provided code which is plotting different things in 2 figures over time using the axes.draw() method.

The plotting of the figures is done with: 

fig_1, axes_1 = plt.subplots(yl,5)  # fig 1
fig_2, axes_2 = plt.subplots(5,1)   # fig 2

and at some point after setting the data plt.draw() is used to redraw the images.

I've tried to add another figure like:

fig_cor, axes_cor = plt.subplots(1,1)

and use 

        axes_cor.imshow(corr_matrix)
        labels = ['P1', 'P2', 'P3', 'P4', 'P5', 'P6', 'P7', 'P8', 'P9', 'P10', 'P11', 'P12', 'P13', 'P14',
                  'P15'] 
        axes_cor.set_xticks(range(0, 15), labels)
        axes_cor.set_yticks(range(0, 15), labels)
        # axes_cor.colorbar()
        plt.draw()

The image is displayed, but my labels are not and when I'm trying to draw the colorbar too I get 'AxesSubplot' object has no attribute 'colorbar'. Can anyone help me solve this?

What I want to display:

enter image description here

What I get:

enter image description here

Upvotes: 1

Views: 2889

Answers (1)

plasmon360
plasmon360

Reputation: 4199

I think in imshow the ticks are created that dont match the label elements. May be there is an elegant solution, but the example below also achieves it.

import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

corr_matrix= np.random.rand(20,20)
fig_cor, axes_cor = plt.subplots(1,1)
fig_cor.set_size_inches(6, 6)

labels = ['P1', 'P2', 'P3', 'P4', 'P5', 'P6', 'P7', 'P8', 'P9', 'P10', 'P11', 'P12', 'P13', 'P14','P15'] 

myimage = axes_cor.imshow(corr_matrix)

plt.colorbar(myimage)

axes_cor.set_xticks(np.arange(0,corr_matrix.shape[0], corr_matrix.shape[0]*1.0/len(labels)))
axes_cor.set_yticks(np.arange(0,corr_matrix.shape[1], corr_matrix.shape[1]*1.0/len(labels)))

axes_cor.set_xticklabels(labels)
axes_cor.set_yticklabels(labels)

plt.draw()

results in

enter image description here

Upvotes: 1

Related Questions