DHARMa
Increasing the validity of statistical analyses with the R package DHARMa. DHARMa uses a simulation-based approach to generate easily interpretable residuals for complex regression models, particularly GLMMs, addressing a widespread difficulty in validating such models’ statistical assumptions. This DFG-funded phase of the project extends the package’s user accessibility and interoperability, adds further diagnostic procedures, and strengthens valid application through testing protocols, in partnership with the R developer community and other package authors. Led by Florian Hartig, with Dr. Melina de Souza Leite and Cosmina Werneke, funded by the Deutsche Forschungsgemeinschaft (DFG, project no. 528747641), 2024–2027.
Research theme: Statistics & machine learning