pdstools.impactanalyzer.Schema¶

Schema definitions for Impact Analyzer experiment data.

These classes mirror the convention used by pdstools.adm.Schema and document the columns expected by ImpactAnalyzer. Required columns are validated in ImpactAnalyzer._validate_ia_data(); optional columns (e.g. those prefixed with Pega_) are populated only by some data sources.

Attributes¶

REQUIRED_IA_COLUMNS

Columns that must be present on the LazyFrame passed to ImpactAnalyzer.

Classes¶

ImpactAnalyzerData

Normalised long-format experiment data consumed by ImpactAnalyzer.

Module Contents¶

class ImpactAnalyzerData¶

Normalised long-format experiment data consumed by ImpactAnalyzer.

Required columns¶

SnapshotTimepl.Datetime | pl.Date: Snapshot date for the row.
ControlGrouppl.Utf8: Control / test group identifier (e.g. "NBA", "PropensityPriority").
Impressionsnumeric: Impression count.
Acceptsnumeric: Accept count.
ValuePerImpressionpl.Float64: Average per-impression value (may be null for PDC data).
Channelpl.Utf8: Channel name (or Channel/Direction for VBD data).

Optional columns¶

Pega_ValueLift, Pega_ValueLiftIntervalpl.Float64: Pre-computed lift values present only in PDC exports.

SnapshotTime¶

ControlGroup¶

Impressions¶

Accepts¶

ValuePerImpression¶

Channel¶

Pega_ValueLift¶

Pega_ValueLiftInterval¶

REQUIRED_IA_COLUMNS: tuple[str, Ellipsis] = ('SnapshotTime', 'ControlGroup', 'Impressions', 'Accepts', 'ValuePerImpression', 'Channel')¶: Columns that must be present on the LazyFrame passed to ImpactAnalyzer.