how to calculate correlation between ten columns with polars

I previously have a large dataframe in pandas and I am having a hard time migrating to Polars.

I used to use the code below to calculate correlation between columns

print(df.corr(numeric_only=True).stack().sort_values(ascending=False).loc[lambda x: x < 1])

and result is like:

how am I supposed to achieve same result with Polars?

many thanks.

You can do it using corr() and unpivot().

(df.corr()
   .with_columns(index = pl.lit(pl.Series(df.columns)))
   .unpivot(index = "index")
   .filter(pl.col("index") != pl.col("variable"))
)

# Output
┌───────┬──────────┬───────────┐
│ index ┆ variable ┆ value     │
│ ---   ┆ ---      ┆ ---       │
│ str   ┆ str      ┆ f64       │
╞═══════╪══════════╪═══════════╡
│ B     ┆ A        ┆ 0.493197  │
│ C     ┆ A        ┆ -0.866325 │
│ D     ┆ A        ┆ -0.493197 │
│ A     ┆ B        ┆ 0.493197  │
│ …     ┆ …        ┆ …         │
│ D     ┆ C        ┆ 0.416025  │
│ A     ┆ D        ┆ -0.493197 │
│ B     ┆ D        ┆ -1.0      │
│ C     ┆ D        ┆ 0.416025  │
└───────┴──────────┴───────────┘