This is important as models that do nor do not converge with the holdout may perform differently with the full dataset. We don't want the whole pipeline to crash every time a model doesn't converge. However, we want to clearly indicate when we cannot report on non-converging models.
This now works need to automatically add note when double dagger features in beta column that those values were from a model with convergence issues. Could be done along with: #64
This is important as models that do nor do not converge with the holdout may perform differently with the full dataset. We don't want the whole pipeline to crash every time a model doesn't converge. However, we want to clearly indicate when we cannot report on non-converging models.