polars.DataFrame.corr

DataFrame.corr(**kwargs: Any) → DataFrame

Return pairwise Pearson product-moment correlation coefficients between columns.

See numpy corrcoef for more information: https://numpy.org/doc/stable/reference/generated/numpy.corrcoef.html

Parameters:

**kwargs

Keyword arguments are passed to numpy corrcoef.

Notes

This functionality requires numpy to be installed.

Examples

>>> df = pl.DataFrame({"foo": [1, 2, 3], "bar": [3, 2, 1], "ham": [7, 8, 9]})
>>> df.corr()
shape: (3, 3)
┌──────┬──────┬──────┐
│ foo  ┆ bar  ┆ ham  │
│ ---  ┆ ---  ┆ ---  │
│ f64  ┆ f64  ┆ f64  │
╞══════╪══════╪══════╡
│ 1.0  ┆ -1.0 ┆ 1.0  │
│ -1.0 ┆ 1.0  ┆ -1.0 │
│ 1.0  ┆ -1.0 ┆ 1.0  │
└──────┴──────┴──────┘