statsmodels

Fit a logistic regression by selecting the GLM family with a logit link, then extract odds ratios for interpretation and run residual diagnostics to validate model assumptions and robustness.

Compare econometric models by evaluating AIC and BIC metrics alongside likelihood-ratio tests, enabling you to select the best-fitting model for your continuous, binary, or count data outcomes.

Yes, you can fit ARIMA and SARIMAX time-series models and evaluate residual behavior by running autocorrelation and heteroskedasticity checks to validate your model assumptions.

Select the GLM family and link function by matching it to your outcome type, such as using Poisson for count data or Gamma for continuous outcomes, ensuring correct data preparation and intercept handling.

Regression residuals may fail normality tests due to outliers or misspecification; interpret influence measures and residual diagnostics to identify problematic data points and validate model robustness.

Yes, the approach supports fitting discrete choice models alongside linear models including OLS, WLS, and GLS, providing inference-ready outputs with confidence intervals and hypothesis testing.