Development build. This page describes main, not a released package. The latest published Lodestar.Preprocessing is 0.1.0 — read its documentation.

Home › Preprocessing › Feature scaling

Undoes Transform, returning values on the original scale.

public double[] InverseTransform(ReadOnlySpan<double> samples)

public CsrMatrix InverseTransform(CsrMatrix samples)

The second overload takes a CsrMatrix and returns a new one storing the same positions, each value multiplied by its column's Scale: a zero stays a zero, so nothing absent becomes stored.

Parameters — samples is the scaled matrix, row-major, with FeatureCount values per row.

Returns — a new array of the same length, back on the input scale.

Exceptions — ArgumentException when samples holds no row, a partial one, or a non-finite value.

The sparse overload throws ArgumentNullException when samples is null, ArgumentException when it holds no row or its column count is not FeatureCount or it stores a non-finite value.

Example — the inverse leaves the unit interval that clipping bounded.

using Lodestar.Preprocessing;

MaxAbsScaler scaler = MaxAbsScaler.Fit([0.0, 10.0], 1, new MaxAbsScalerOptions { Clip = true });

double back = scaler.InverseTransform([5.0])[0];  // => 50

Remarks — it never clips, even when the scaler does, for the reason MinMaxScaler.InverseTransform gives: a clipped value has lost what it was.

Applies to — net10.0, netstandard2.0.

See also — MaxAbsScaler.Transform, MaxAbsScaler.