spacr.row_exclusions

General row-exclusion rules shared by UMAP and its parameter search.

Functions

exclude_matching_rows(→ tuple[Any, list[str]])	Drop rows matching any configured column/value rule.
normalize_row_exclusions(→ dict[str, list[Any]])	Return {column: [values...]} from settings or CSV text.

Module Contents

spacr.row_exclusions.exclude_matching_rows(frame, rules: Any) → tuple[Any, list[str]]

Drop rows matching any configured column/value rule.

Parameters:

• frame – table whose rows should be filtered.

• rules – mapping or serialized row-exclusion rules.

Values are compared both in their native dtype and as stripped strings. This lets a value selected from SQLite text match the equivalent pandas numeric value without changing identifiers such as "001".

Returns:

(filtered_frame, notes). Each note names the column, values, and number of rows removed.

Raises:

ValueError – for unknown columns or rules that remove every row.

spacr.row_exclusions.normalize_row_exclusions(value: Any) → dict[str, list[Any]]

Return {column: [values...]} from settings or CSV text.

Parameters:

value – mapping or serialized mapping of columns to excluded values.

None and an empty value mean no exclusions. A scalar value is accepted as a one-item list so hand-written settings remain convenient.

Returns:

stripped column names mapped to ordered, deduplicated values.

Raises:

ValueError – when a nonempty value cannot be parsed as a mapping.