Repository navigation
Tags
Tags
Reorganize repo: data/, pipeline/, scripts/ directories - data/: estimation output CSVs (archived by Zenodo) - pipeline/: SAS data build + HDF5 conversion - scripts/: job submission wrappers (SGE, Slurm) + utilities - Remove dead files: data.sas (superseded), index.html, README.org Generated with [Claude Code](https://claude.ai/code) via [Happy](https://happy.engineering) Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Happy <yesreply@happy.engineering>
Vectorize DY model loglik (precompute lfact, remove retry loops) Same pattern as GPIN optimization: precompute data constants, np.logaddexp chain, eliminate while-loop retries, numpy upfront. EEOW left as-is (loglik/fit are unimplemented stubs). Generated with [Claude Code](https://claude.ai/code) via [Happy](https://happy.engineering) Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Happy <yesreply@happy.engineering>