Stata 17 offers new features that enhance our lasso suite. We added lasso treatment effects, which provide causal inference with the robustness properties of the treatment-effects estimators while controlling for high-dimensional covariates by using lasso. We have new options that allow you to account for clustered data in your analysis. Also, you may now use the BIC criterion to select the lasso-penalty parameter. Read more about these features below.

• Lasso for treatment effects
• Lasso with clustered data
• BIC for lasso penalty selection

Additional resources

Learn more about Stata's lasso features.

Read more about lasso in the Stata Lasso Reference Manual.

Learn about treatment effects in lasso in [TE] telasso.

View all the new features