Correlation Matrix

heat_map() is a natural fit for correlation matrices — the symmetric structure and [-1, 1] value range map perfectly onto a divergent colormap, making strong positive and negative correlations immediately visible.

This example uses Fisher's Iris dataset: 150 samples across 3 species, with 4 numeric features (sepal/petal length and width).

import polars as pl
import polarise
from polarise.datasets import get_iris_data
df = get_iris_data()

corr_df = df[:, :-1].corr(label='features').with_columns(
    pl.col(pl.Float64).round(2)
)

---

Correlation matrix

{ cmap="CET_L19" · built-in or colorcet }

(corr_df.style()
        .heat_map(exclude='features', cmap='CET_L19')
        .footnote('source : UCI Machine Learning Repository — Fisher\'s Iris dataset')
        .fashion_grid()
        .show()
 )

features	sepal_length	sepal_width	petal_length	petal_width
sepal_length	1.0	-0.12	0.87	0.82
sepal_width	-0.12	1.0	-0.43	-0.37
petal_length	0.87	-0.43	1.0	0.96
petal_width	0.82	-0.37	0.96	1.0

source: UCI Machine Learning Repository — Fisher's Iris dataset