statsmodels

Run robust inference in Python by fitting OLS regression models and applying residual diagnostics, heteroskedasticity checks, and robust standard errors to validate assumptions and quantify uncertainty.

Fit time series models like ARIMA and SARIMAX to generate forecasts with confidence intervals, applying model comparison metrics like AIC and BIC alongside interval summaries for rigorous time series forecasting.

Model discrete outcomes and count data for econometrics by fitting GLMs like Binomial or Poisson, running core diagnostics, and returning coefficient summaries with uncertainty measures for publication-ready results.

This approach supports influence analysis and outlier detection for statistical models by applying residual diagnostics and influence measures to identify impactful observations during model fitting and inference checks.

Compare statistical models using likelihood-based metrics by calculating AIC and BIC values across fitted regression or discrete choice models to evaluate fit and select the most appropriate specification.