Crate tdigests

T-Digest Rust Library

This library provides an implementation of the t-digest algorithm in Rust. The t-digest algorithm is a data structure for accurate on-line accumulation of rank-based statistics such as quantiles and trimmed means.

Features

• Efficient computation of quantiles over streaming data.
• Suitable for large datasets and distributed systems.
• Supports merging of t-digests from different data partitions.

Example

use tdigests::TDigest;

// Create a t-digest from a list of values
let values = vec![10.0, 20.0, 30.0];
let mut digest = TDigest::from_values(values);

// Compress the centroids with a maximum of 100 centroids
digest.compress(100);

// Estimate quantiles
let median = digest.estimate_quantile(0.5);
assert_eq!(median, 20.0);

// Compute the relative rank of a value
let rank = digest.estimate_rank(25.0);
assert_eq!(rank, 0.75);

Structs

Centroid

Represents a centroid in the t-digest, which holds a mean value and an associated weight.

TDigest

The main struct representing a t-digest.