Development build. This page describes main, not a released package. The latest published Lodestar.Text is 0.7.0 — read its documentation.

Home › Text › Vectorization

The term each column stands for, in column order.

public IReadOnlyList<string> GetFeatureNames()

Returns — IReadOnlyList<string> of length ColumnCount, sorted by code point, where index i is the term counted by column i.

Exceptions — InvalidOperationException when nothing has been fitted yet.

Example — the vocabulary is sorted, so the order depends on the terms and not the documents.

using Lodestar.Text.Vectorization;

var cv = new CountVectorizer();
cv.Fit(["the cat eats", "the dog eats", "the cat and the dog"]);

IReadOnlyList<string> names = cv.GetFeatureNames();

string first = names[0];   // => and
string last = names[4];    // => the

Remarks — this is what makes a count matrix readable, and it is the thing HashingVectorizer cannot offer: hashing throws the vocabulary away, so there is no name to return and no method to return it. Choosing that vectorizer is choosing to give this up.

The list is a read-only view of the vectorizer's own vocabulary, made at the first call and not copied, so reading it allocates nothing after that; a cast back to an array fails rather than editing the fitted model under a later Save (#1670).

Applies to — net10.0, netstandard2.0.

See also — CountVectorizer, CsrMatrix.