pdstools.prediction.Schema¶

Schema definitions for Prediction Studio data.

Mirrors the layout of pdstools.adm.Schema — each class is a plain attribute container mapping column names to their target Polars dtypes. Used by pdstools.prediction.Prediction’s internal validators to coerce raw input into a consistent shape.

Classes¶

`PredictionSnapshot`	Raw column dtypes for the `PR_DATA_DM_SNAPSHOTS` Pega export.
`PredictionData`	Processed/output column dtypes for `Prediction.predictions`.

Module Contents¶

class PredictionSnapshot¶

Raw column dtypes for the PR_DATA_DM_SNAPSHOTS Pega export.

These are the columns produced by Pega’s Prediction Studio dataset export (or equivalent PDC payload) before any pdstools renaming or pivoting is applied.

pySnapShotTime¶

pyModelId¶

pyModelType¶

pySnapshotType¶

pyDataUsage¶

pyPositives¶

pyNegatives¶

pyCount¶

pyValue¶

pyName¶

class PredictionData¶

Processed/output column dtypes for Prediction.predictions.

The shape produced after Prediction’s validator pivots raw rows into per-(pyModelId, SnapshotTime) records with Test/Control/NBA suffixes.

pyModelId¶

SnapshotTime¶

Class¶

ModelName¶

Performance¶

Positives¶

Negatives¶

ResponseCount¶

Positives_Test¶

Positives_Control¶

Positives_NBA¶

Negatives_Test¶

Negatives_Control¶

Negatives_NBA¶

ResponseCount_Test¶

ResponseCount_Control¶

ResponseCount_NBA¶

CTR¶

CTR_Test¶

CTR_Control¶

CTR_NBA¶

CTR_Lift¶

isValidPrediction¶