Module diagnostics 

Statistical diagnostic tests for linear regression assumptions.

This module provides a comprehensive suite of diagnostic tests to validate the assumptions of ordinary least squares (OLS) regression. Each test is implemented in its own file for easier maintenance.

Available Tests

Linearity Tests

• Rainbow Test (rainbow.rs) - Tests whether the relationship between predictors and response is linear
• Harvey-Collier Test (harvey_collier.rs) - Tests for functional form misspecification using recursive residuals

Heteroscedasticity Tests

• Breusch-Pagan Test (breusch_pagan.rs) - Tests for constant variance of residuals (studentized/Koenker variant)
• White Test (white.rs) - More general test for heteroscedasticity that does not assume a specific form

Normality Tests

• Jarque-Bera Test (jarque_bera.rs) - Tests normality using skewness and kurtosis
• Shapiro-Wilk Test (shapiro_wilk.rs) - Powerful normality test for small to moderate samples (n ≤ 5000)
• Anderson-Darling Test (anderson_darling.rs) - Tail-sensitive test for normality

Autocorrelation Tests

• Durbin-Watson Test (durbin_watson.rs) - Tests for first-order autocorrelation in residuals

Influence Measures

• Cook’s Distance (cooks_distance.rs) - Identifies influential observations that may affect regression results

Structs

CooksDistanceResult

Result of Cook’s distance analysis.

DiagnosticTestResult

Result of a diagnostic test.

DurbinWatsonResult

Result of the Durbin-Watson test

RainbowSingleResult

Result of the Rainbow test for a single method (R or Python).

RainbowTestOutput

Result of the Rainbow test supporting both R and Python variants.

WhiteSingleResult

Result of the White test for a single method (R or Python).

WhiteTestOutput

Result of the White test supporting both R and Python variants.

Enums

RainbowMethod

Rainbow test implementation method.

WhiteMethod

White test implementation method.

Functions

anderson_darling_test

Performs the Anderson-Darling test for normality of residuals.

anderson_darling_test_raw

Applies the Anderson-Darling test directly to a sample of values.

breusch_pagan_test

Performs the Breusch-Pagan test for heteroscedasticity.

cooks_distance_test

Computes Cook’s distance for identifying influential observations.

durbin_watson_test

Durbin-Watson test for first-order autocorrelation in residuals.

f_p_value

Computes a p-value from an F-statistic.

harvey_collier_test

Performs the Harvey-Collier test for linearity (functional form).

jarque_bera_test

Performs the Jarque-Bera test for normality of residuals.

python_white_method

Performs the White test for heteroscedasticity using Python’s method.

r_white_method

Performs the White test for heteroscedasticity using R’s method.

rainbow_test

Rainbow test with method selection (R, Python, or Both).

shapiro_wilk_test

Performs the Shapiro-Wilk test for normality of residuals.

shapiro_wilk_test_raw

Applies the Shapiro-Wilk test directly to a sample of values.

two_tailed_p_value

Computes a two-tailed p-value from a t-statistic.

validate_regression_data

Validates regression input data for dimensions and finite values.

white_test

Performs the White test for heteroscedasticity.