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

Home › Preprocessing › Encoding and imputation

Fills the missing values of a row-major sample matrix.

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

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

Returns — a new array of the same length, with every NaN replaced.

Exceptions — ArgumentException when samples holds no row, a partial one, or an infinity.

Example — the statistic is the fit's, not the transform's.

using Lodestar.Preprocessing;

double[] fitted = [1.0, 2.0, 3.0];

SimpleImputer imputer = SimpleImputer.Fit(fitted, featureCount: 1);

// The mean of 1, 2 and 3 — the row being filled has no say in it.
string filled = string.Join(",", imputer.Transform([double.NaN, 99.0]));  // => 2,99

Remarks — the width never changes, which is the divergence Fit explains: the reference's default returns one column fewer when a feature was empty at fit time.

Never writes to the input.

Applies to — net10.0, netstandard2.0.

See also — SimpleImputer, SimpleImputerOptions.