solow 0.7.3

The comprehensive statistics and machine learning stack for Rust. Regression, generalized linear models, discrete choice, time series, state space, survival analysis, mixed effects, Bayesian inference, clustering, tree ensembles, SVM, neural networks, kernel methods, dimensionality reduction, and change point detection. 57 focused crates, memory safe, pure Rust.
Documentation

solow

The comprehensive statistics and machine learning stack for Rust. 57 focused crates. Memory safe. Pure Rust. Deterministic.

Install

[dependencies]
solow = "0.7"

Quick start

use solow::prelude::*;
use ndarray::array;

let y = array![2.5, 3.4, 4.7, 5.1, 6.5];
let x = array![[1.0, 1.0], [1.0, 2.0], [1.0, 3.0], [1.0, 4.0], [1.0, 5.0]];
let res = LinearModel::ols(y, x).unwrap().fit().unwrap();
assert!(res.rsquared > 0.98);

What is covered

Group Modules
Linear models regression, glm, discrete, robust
Time series tsa, statespace, var, regime
Panel and survival mixed, gee, duration
Multivariate multivariate, cross_decomposition, covariance
Machine learning svm, tree, ensemble, neural, naive_bayes, neighbors, discriminant, cluster
Dimensionality reduction decomposition, manifold, kernel_approx
Semi, multi, probability semi_supervised, multi, calibration, gp
Model selection and metrics cv, metrics, pipeline, feature_selection
Preprocessing and data preprocessing, text, impute, datasets
Bayesian and formulas bayes, emplike, copula, formula, fit
Statistics stats, nonparametric, gam, othermod
Distributions and graphics distributions, viz, graphics, summary

Every subcrate is reachable at solow::<module>.

Beyond the classical stack

Capabilities that few libraries expose as first-class modules.

  • Change-point detection. cusum, pelt (Killick, Fearnhead, Eckley 2012), binary_segmentation.
  • Volatility. Garch11 with iterated multi-step variance forecast.
  • Extreme value analysis. Gev, Gpd with return_level(T) and peaks-over-threshold fit.
  • Effect sizes. cohens_d, hedges_g, glass_delta, eta_squared, omega_squared, cliffs_delta, cramers_v.
  • Meta-analysis. meta_fixed_effect, meta_random_effects with Cochran Q, I squared, tau squared.
  • Control charts. ewma and two-sided cusum.
  • Block bootstrap. Moving, circular, stationary variants for time series.

Correctness

Every deterministic estimator is cross-verified against committed golden reference fixtures. Closed-form solvers match bit-wise to 1e-10. Iterative solvers match parameters to 1e-6 or predictions to 5e-2. Every crate lives under #![forbid(unsafe_code)].

Full workspace

The workspace ships 57 focused crates. Depend on the umbrella for the full surface or pick individual crates for a leaner build.

[dependencies]
solow-regression = "0.7"
solow-metrics    = "0.7"
solow-cv         = "0.7"

See the full workspace list and the documentation site for per-module walkthroughs.

License

BSD-3-Clause. Copyright (c) 2026, Murat Ova (Stochastic Minds).