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, or copied unchanged when the scaler does not scale: a zero stays a zero, so nothing absent becomes stored.

Parameters — samples is the transformed 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, and InvalidOperationException when the scaler centres — fit it with WithCentring = false, since subtracting a centre would make every absent zero a stored value.

Example — the round trip returns what went in.

using Lodestar.Preprocessing;

double[] samples = [1.0, 2.0, 3.0, 4.0, 5000.0];

RobustScaler scaler = RobustScaler.Fit(samples, featureCount: 1);

double back = scaler.InverseTransform(scaler.Transform(samples))[4];  // => 5000

Remarks — multiplies then adds, the reverse of the transform's order. Exact up to floating-point rounding, except for a feature whose range was floored to 1: that step threw the spread away rather than recording it.

Applies to — net10.0, netstandard2.0.

See also — RobustScaler.Transform, RobustScaler.