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

CountVectorizer.Save

Write a fitted vectorizer out, so the vocabulary survives the process.

public void Save(Stream destination)
public void Save(string path)

Parameters — destination is a writable stream, left open for the caller to dispose; path is a file to create or overwrite.

Exceptions — InvalidOperationException when nothing has been fitted yet, refused before the path is opened. ArgumentException, the JSON writer's own, as 0.7.0 raised it, when a vocabulary term, the token pattern or a stop word is beyond what the writer can write — about 166 million characters, fewer where it escapes — refused before a byte is written, once the writer has accepted the stream or once the file is open, so a path opening refuses is refused ahead of it (#1618, #1630, #1646); ArgumentException too when the stream cannot be written to, refused once the vectorizer is known fitted, before its strings are checked (#1641). IndexOutOfRangeException, the writer's own as well, for a string whose escaped form passes its buffer, as 0.7.0 raised it. ArgumentNullException for a null stream or path. IOException from the stream or file system.

Example — round-tripping through memory.

using Lodestar.Text.Vectorization;

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

using var buffer = new MemoryStream();
cv.Save(buffer);
buffer.Position = 0;

CountVectorizer restored = CountVectorizer.Load(buffer);
int columns = restored.Transform(["the cat"]).ColumnCount;  // => 4

Remarks — what is written is the fit: the vocabulary and the options that produced it. A vectorizer restored from it counts a corpus exactly as the original would, which is the point — the alternative is refitting on training data that may no longer be around.

Saving before fitting throws rather than writing an empty vocabulary, because a file that loads into a vectorizer which drops every term is worse than no file.

The stream overload leaves destination open. That is deliberate: it lets a vectorizer be one part of a larger archive.

The write is not bounded the way a load is. A vocabulary past a default of ArtifactLoadOptions — a token over 1024 characters, more than a million features — saves, and loads only with that bound raised.

Applies to — net10.0, netstandard2.0.

See also — CountVectorizer.Load, CountVectorizer.SaveAsync.