linreg-core 0.8.1

Lightweight regression library (OLS, Ridge, Lasso, Elastic Net, WLS, LOESS, Polynomial) with 14 diagnostic tests, cross validation, and prediction intervals. Pure Rust - no external math dependencies. WASM, Python, FFI, and Excel XLL bindings.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

{
  "test_name": "Breusch-Godfrey Test (Python - statsmodels)",
  "dataset": "iris",
  "formula": "Sepal.Length ~ Sepal.Width + Petal.Length + Petal.Width + Species",
  "order": 1,
  "statistic": 17.017815146966342,
  "p_value": 3.7030737947018e-05,
  "test_type": "Chisq",
  "df": [
    1.0
  ],
  "f_statistic": 17.98259884698945,
  "f_p_value": 3.26164527240011e-05,
  "f_df": [
    1.0,
    223.0
  ],
  "passed": false,
  "interpretation": "Significant serial correlation detected at order <= 1",
  "description": "Tests for serial correlation up to order 1. The LM (Chi-squared) statistic is asymptotically distributed as chi-squared with 1 degrees of freedom. The F statistic provides a finite-sample correction. Uses statsmodels.stats.diagnostic.acorr_breusch_godfrey."
}