Hi Astha,

Thanks for showing your interest in this article.

This pictorial representation represents the correlation between the same column itself like num_preg with num_preg, etc. To understand better we have the following patterns as per the plotted graph, just follow the yellow color in the graph:

num_preg->num_preg

glucose_conc->glucose_conc

diastolic_bp->diastolic_bp

thickness->thickness

thickness->skin (correlation doesn’t match as columns differ)

insulin->insulin

bmi->bmi

diab_pred->diab_pred

age->age

skin->thickness (correlation doesn’t match as columns differ)

skin->skin

diabetes->diabetes

Hence, we talked about removing either skin or thickness column.

The same applies to the table:

It is just the numerical representation of the graph. And the values have been calculated by .corr() function which is the part of pandas library.