pub fn bic_python(log_likelihood: f64, n_coef: usize, n_obs: usize) -> f64Expand description
Computes BIC using Python/statsmodels convention.
BIC = k*ln(n) - 2logL
where k is the number of coefficients (NOT including variance parameter). This matches Python’s statsmodels OLS.bic behavior.
§Arguments
log_likelihood- Log-likelihood of the modeln_coef- Number of coefficients (including intercept)n_obs- Number of observations
§Example
let bic_value = bic_python(-150.5, 3, 100); // 3 coefficients, 100 obs