Implements methods for covariate-adjusted local likelihood estimation and inference in the sharp regression discontinuity design with an unordered multinomial outcome. A comprehensive description of the model, estimation framework, bandwidth selection and inference procedures is provided in the accompanying paper, "Nonlinear Regression Discontinuity Designs with Covariates."

Implements local likelihood estimation and inference procedures for (unordered) multinomial treatment effects in both the sharp and fuzzy regression discontinuity designs. The model and methods for the fuzzy design are introduced in "Fuzzy Regression Discontinuity Designs with Multinomial Outcomes." The paper and package also include several improvements to the bandwidth selection and inference procedures for the sharp design, building on the original framework developed in Xu (2017), "Regression Discontinuity with Categorical Outcomes," Journal of Econometrics.