Explode

Bases: Transformer

              flowchart TD
              matchlab.transformers.explode.Explode[Explode]
              matchlab.transformers.base.Transformer[Transformer]

                              matchlab.transformers.base.Transformer --> matchlab.transformers.explode.Explode
                


              click matchlab.transformers.explode.Explode href "" "matchlab.transformers.explode.Explode"
              click matchlab.transformers.base.Transformer href "" "matchlab.transformers.base.Transformer"
            

Give every combination of each id's non-null column values its own row.

Group collapses each id to one row. Explode goes the other way, giving the true cross product a list aggregate can't express. Every non-null value in one column pairs with every non-null value in every other column, for the same id. Values are deduplicated per column first, so a value repeated across records counts once. A column that is all null for one id still keeps that row, with null there, instead of dropping it.

An id with several populated columns can generate many rows fast, so max_combinations caps it and raises, naming the id, rather than silently building an enormous table.

Methods:

• apply –

  Cross-join each id's columns to every combination of their values.

Attributes:

• version (int) –
• max_combinations (int) –
• model_config –

version: int = 1

max_combinations: int = Field(default=100, description='Reject a group whose cross product would exceed this many rows.')

model_config = ConfigDict(extra='forbid', frozen=True)

apply(data: DataFrame) -> DataFrame