Lodestar.Preprocessing 0.2.0. This page is frozen at that release. Read the current documentation for what main says now. A link to a decision or a migration page follows main, and leaves the archive.

Home › Feature scaling

StandardScalerOptions

Which of the two standardisation steps StandardScaler applies.

public sealed record StandardScalerOptions

Properties — WithMean centres each feature on its mean, WithStd divides it by its standard deviation. Both default to true, as sklearn.preprocessing.StandardScaler's with_mean and with_std do.

Example — scale without centring, which is what a caller keeping sparsity asks for.

using Lodestar.Preprocessing;

double[] samples = [1.0, 10.0, 2.0, 10.0, 4.0, 10.0];

StandardScaler scaler = StandardScaler.Fit(
    samples, featureCount: 2, new StandardScalerOptions { WithMean = false });

// The mean is still fitted — turning centring off does not stop it being computed.
double mean = scaler.Mean![0];              // => 2.3333333333333335

// But it is not subtracted: the first value keeps its own magnitude.
double first = scaler.Transform(samples)[0]; // => 0.8017837257372732

Remarks — the pair does more than switch two lines of arithmetic: it decides which fitted statistics exist at all, and the mapping is not symmetric. StandardScaler carries the table.

WithMean = false is the option a caller reaches for when centring would be wrong rather than merely unwanted — subtracting a mean from a matrix whose zeros are meaningful turns every zero into a non-zero, which is why scikit-learn refuses to centre a sparse matrix at all.

Applies to — net10.0, netstandard2.0.

See also — StandardScaler, StandardScaler.Fit.