seaborn clustermap does not rearrange annotation data

Question

I am using seaborn to plot a clustermap from a matrix, and I want the values of the matrix to be the annotations of the clustermap. However, while the clustermap rearranges rows and columns of the data matrix to form clusters, it does not do so for the annotation matrix.

Here is my code:

data = np.random.rand(10,5)
print (data)

 [[0.62906695 0.41046601 0.04774079 0.98012573 0.79686483]
 [0.1618218  0.70365808 0.53644302 0.68291708 0.95152119]
 [0.901431   0.19378611 0.33176554 0.00219459 0.92936903]
 [0.1575622  0.37749714 0.88975381 0.16945422 0.23790588]
 [0.06837024 0.49287138 0.60866541 0.80563271 0.38924653]
 [0.13671484 0.11410341 0.14604541 0.49899899 0.82850949]
 [0.90237875 0.23954254 0.41614074 0.11811778 0.26017433]
 [0.27881206 0.09321617 0.62520404 0.10796313 0.12266986]
 [0.27606845 0.8479972  0.86467503 0.19660883 0.11909075]
 [0.64211049 0.61860723 0.70426254 0.24465052 0.17614033]]

df = pd.DataFrame(data, columns=['A','B','C','D','E']).corr()
df

          A         B         C         D         E
A  1.000000 -0.229103 -0.353145 -0.360782  0.084730
B -0.229103  1.000000  0.498165  0.256117 -0.162689
C -0.353145  0.498165  1.000000 -0.443952 -0.727170
D -0.360782  0.256117 -0.443952  1.000000  0.448025
E  0.084730 -0.162689 -0.727170  0.448025  1.000000


sns.clustermap(data=df, annot=df)

As you can see in the figure, the annotations are the same with the original df and the values have not been arranged according to the clusters in the figure.

Is there any way I can do this correctly?

Answer 1

You are manually setting the annotations, just use

sns.clustermap(data=df, annot=True)