[FIXED] create seaborn heatmap from multiple columns

Issue

I have a dataframe as such

col1    col2    col3    col4    col5   col6    col7     col8     col9      col10

2       12      8        3      16      9       5        13       11        14
3       14      7        9      18      12      14       14       13        13
5       15      10       5      8       10      18       9         8         5

I want the output to produce a heatmap where col1-col5 are compared vs col6-col10

col1

col2

col3

col4

col5

       col6    col7  col8  col9    col10

How can I achieve such a structure?

Solution

The following example code shows how you can correlate each of the last 5 columns with each of the first 5 columns, and show the result as a heatmap.

import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
import pandas as pd

# create some test data
df = pd.DataFrame(np.random.rand(3, 10), columns=[f'col{i}' for i in range(1, 11)])
# find the correlations between the last 5 columns with each of the 5 first columns
corr = df[df.columns[5:]].apply(lambda x: df[df.columns[:5]].corrwith(x))
# show the correlations as a heatmap
ax = sns.heatmap(corr, annot=True, cmap='RdYlBu', vmin=-1, vmax=1)
ax.tick_params(labelrotation=0)
plt.tight_layout()
plt.show()

Answered By - JohanC