Open felix-clark opened 2 years ago
A common feature many statisticians would want is the ability to easily get a p-value for a model or for a particular covariate. Right now p-values can typically be calculated from Z-test or likelihood test values, but the user would need to call special mathematical functions and the process would be confusing.
A crate like
special-fun
would enable some calculations like CDFs and quantile residuals. Perhaps this should be gated behind an optional feature.