- Conditional likelihood (I already have a branch with first order approximation, but I want to implement second and third order, and also the case without approximation for backward nonlinear models).
- Method of Moments (only the documentation is still missing).
- Enable estimating models with unanticipated structural break.
- Bayesian methods of moments (à la Christiano et al.).
- Pruning at higher order (#1643)
- Report of accuracy errors.
- Complete interface for extended path.
dseries & dates classes:
- Interface to DBnomics.
- Add id to each variable in dseries objects.
- Allow mixed frequencies or create a new object with dseries of different frequencies.
move to Dragonfly for parallel execution