Member Junction
    Preparing search index...

    The unit version a model was trained on, the typed feature columns its matrix was assembled from, the kept levels of each categorical column and the Training Snapshot its rows came from.

    interface UnitModelLineage {
        categoryLevels: Record<string, string[]>;
        featureSpec: FeatureColumnSpec[];
        featureSpecVersion: number;
        labelVersion: number;
        predictionUnitId: string;
        predictionUnitVersion: number;
        snapshotContentHash: string;
        trainingSnapshotId: string;
        unitRowEntity: string;
    }
    Index

    Properties

    categoryLevels: Record<string, string[]>

    The kept levels of each categorical column, by column name; scoring folds every other value into the fold level.

    featureSpec: FeatureColumnSpec[]

    The typed columns the matrix was assembled from, in order.

    featureSpecVersion: number
    labelVersion: number
    predictionUnitId: string
    predictionUnitVersion: number
    snapshotContentHash: string

    That snapshot's content hash, so the lineage names the exact rows even if the row were ever re-pointed.

    trainingSnapshotId: string

    The MJ: ML Training Snapshots row the model was trained from.

    unitRowEntity: string

    The entity whose records the unit rows are (the entity a scoring scope must hold).