eval_aligned

pydiverse.transform.eval_aligned(

val: ColExpr | Series | Series,

with_: Table | Col | None = None,

) → EvalAligned

Allows to evaluate a column expression containing columns from different tables and to use polars / pandas Series in column expressions.

Parameters:

• val – The expression or polars / pandas Series to be aligned.

• with – The table or column to align with.

Examples

Usage of a polars Series in a column expressions (the same works for pandas Series):

>>> import polars as pl
>>> t = pdt.Table(
...     {
...         "a": [1, 2, 3, 4],
...         "b": [2, 5, 16, 3],
...     },
...     name="t",
... )
>>> s = pl.Series([9, 5, 4, 1])
>>> t >> mutate(c=eval_aligned(t.a + s)) >> show()
Table `t` (backend: polars)
shape: (4, 3)
┌─────┬─────┬─────┐
│ a   ┆ b   ┆ c   │
│ --- ┆ --- ┆ --- │
│ i64 ┆ i64 ┆ i64 │
╞═════╪═════╪═════╡
│ 1   ┆ 2   ┆ 10  │
│ 2   ┆ 5   ┆ 7   │
│ 3   ┆ 16  ┆ 7   │
│ 4   ┆ 3   ┆ 5   │
└─────┴─────┴─────┘

Expression containing columns from different tables:

>>> t1 = pdt.Table({"a": [1, 2, 3, 4]}, name="t1")
>>> t2 = pdt.Table({"a": [5, 3, 1, 3]}, name="t2")
>>> t1 >> mutate(c=eval_aligned(t1.a + t2.a, with_=t1)) >> show()
Table `t1` (backend: polars)
shape: (4, 2)
┌─────┬─────┐
│ a   ┆ c   │
│ --- ┆ --- │
│ i64 ┆ i64 │
╞═════╪═════╡
│ 1   ┆ 6   │
│ 2   ┆ 5   │
│ 3   ┆ 4   │
│ 4   ┆ 7   │
└─────┴─────┘