Text 0.3.1 osa normalizedsimilarity

Lodestar.Text 0.3.1. 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.

Osa.NormalizedSimilarity

1 - NormalizedDistance: 1 when the two are identical, 0 when nothing survives.

public static double NormalizedSimilarity(ReadOnlySpan<char> a, ReadOnlySpan<char> b, TextElement element = TextElement.Utf16Unit)

Parameters — a and b are the two strings to compare; element says what counts as one character, exactly as it does for NormalizedDistance, which this is computed from.

Returns — double in [0, 1], larger meaning more alike. Two empty inputs give 1.

Example — the same swap, read as a score.

using Lodestar.Text.Distances;

double s = Osa.NormalizedSimilarity("abcd", "acbd");   // => 0.75

Remarks — the member to rank on when transpositions should be forgiven and a bigger number should mean a better match. It matches OSA.normalized_similarity in rapidfuzz.

Two traps, both inherited. Two empty inputs return 1. And short inputs make the scale coarse: with max(len(a), len(b)) as the divisor, a four-character pair can only ever score 0, 0.25, 0.5, 0.75 or 1, so a threshold like 0.8 is really a threshold of 1 for anything that short.

Applies to — net10.0, netstandard2.0.

See also — Osa.NormalizedDistance, DamerauLevenshtein.NormalizedSimilarity, Levenshtein.NormalizedSimilarity, the Python equivalence table.