Context
PR #11 migrated the seasonal model from dummy-variable regression to Kalman state-space. As a result, build_seasonal_design() and the seasonal_x field are now dead code.
What to remove
state_space.rs: build_seasonal_design() (L159-177), seasonal_x field, num_seasonal_covariates(), seasonal_covariates()state_space.rs: 2 unit tests for build_seasonal_designsampler.rs: all seasonal_beta / k_seasonal arguments and their usage (~80 lines)
Why
When nseasons > 1, seasonal_x = vec![] always (state-space path). When nseasons <= 1, build_seasonal_design returns empty vec. The field is always empty in practice.
Context
PR #11 migrated the seasonal model from dummy-variable regression to Kalman state-space. As a result,
build_seasonal_design()and theseasonal_xfield are now dead code.What to remove
state_space.rs:build_seasonal_design()(L159-177),seasonal_xfield,num_seasonal_covariates(),seasonal_covariates()state_space.rs: 2 unit tests forbuild_seasonal_designsampler.rs: allseasonal_beta/k_seasonalarguments and their usage (~80 lines)Why
When
nseasons > 1,seasonal_x = vec![]always (state-space path). Whennseasons <= 1,build_seasonal_designreturns empty vec. The field is always empty in practice.