Member Junction
    Preparing search index...

    Seeded, stratified locked-holdout carve. Pure; no I/O.

    Index

    Constructors

    Methods

    • The split a Training Snapshot's row manifest records: the development rows are the training portion and the locked-holdout rows the holdout, each in matrix order, with every snapshot column. Nothing is re-carved, so the counts and the holdout class counts describe exactly the rows the manifest names.

      Parameters

      • content: SnapshotContent

        the snapshot content (matrix row i is manifest line i)

      • labelColumn: string

        the label column the holdout class counts are read from

      • problemType: ProblemType

        classification counts the holdout rows per label; regression stores no class counts

      Returns LockedHoldoutSplit

    • How many rows a fraction holds out of total: at least one when the fraction is positive, never every row (one row always stays for training), and none when there is at most one row.

      Parameters

      • total: number
      • fraction: number

      Returns number

    • Mulberry32: a small, fast 32-bit seeded generator returning floats in [0, 1). Local so the carve has no dependency whose algorithm could change under it — the same seed must give the same holdout across releases.

      Parameters

      • seed: number

      Returns () => number