带有 48 个变量的 Seaborn 热图

Question

I have a dataset for which I have filtered down my the possible attributes to be used for model building , they are still coming out to be 48 variables. When I try to draw a corelation map using seaborn heatmap , the values are overwritten and not clear.

Code used

    cor = train[cols].corr()
    #fig,ax = plt.subplots()
    sns.heatmap(cor,annot=True)
    plt.show()

How can I make this readable ,have tried figsize but still not working. I am using Pycharm

Answer 1

You can change the fontsize of the annotation by passing annot_kws to the function:

sns.heatmap(np.arange(100).reshape(10,10), annot=True)

sns.heatmap(np.arange(100).reshape(10,10), annot=True, annot_kws={'fontsize':5})