Development build. This page describes main, not a released package. The latest published Lodestar.Preprocessing is 0.2.0 — read its documentation.

Home › Preprocessing › Encoding and imputation

What OneHotEncoder does with a value the fit never saw.

public enum UnknownCategory

Fields — Refuse raises naming the feature, and Ignore encodes the value as all zeros — handle_unknown="error" and "ignore", with Refuse the default as "error" is. Infrequent encodes it in its feature's infrequent column, or as all zeros where the feature has none — "infrequent_if_exist"; the reference's "warn" does the same with a warning, which this does not raise.

Example — the same value under both settings.

using Lodestar.Preprocessing;

OneHotEncoder<string> lenient = Encoders.OneHot(
    ["a", "b"], 1, new OneHotEncoderOptions { Unknown = UnknownCategory.Ignore });

string unknown = string.Join(",", lenient.Transform(["zzz"]));  // => 0,0

Remarks — refusing is the default because the alternative is silent: an unknown category becomes a row of zeros that a model reads as a category of its own, and no exception marks the point where the data stopped matching the fit.

Applies to — net10.0, netstandard2.0.

See also — OneHotEncoderOptions, OneHotEncoder.Transform.