crabbymetrics

crabbymetrics is a Rust-backed econometrics library with a compact Python API. This homepage is the entry point for the Quarto docs site in docs/.

Main Reference

• API reference: verified public surface, library sitrep, summary schemas, and runtime smoke checks.

Highlights

• PCA and KernelBasis let you build richer right-hand sides before fitting the regression estimators.

Examples

• OLS: baseline linear regression with HC1 standard errors.
• Fixed Effects OLS: partial out one-way or multi-way categorical fixed effects with within, then estimate slopes without an intercept.
• Synthetic Control: simplex-constrained donor weighting for treated-versus-donor panel matching under latent-factor drift.
• PCA And Kernel Basis: linear low-rank compression for factor-style designs, plus nonlinear kernel features on a swiss roll.
• Richer Regression: use KernelBasis and PCA to build a nonlinear right-hand side that beats raw OLS on a held-out test set.
• ElasticNet: regularized linear regression.
• Logit: binary logistic regression.
• Multinomial Logit: multiclass classification.
• Poisson: count regression.
• TwoSLS: instrumental variables regression.
• FTRL: online-style binary classification.
• MEstimator Poisson: callback-driven estimation matched against built-in Poisson.

Notes

• api.qmd remains the main documentation page and renders to api.html.
• The site is a Quarto website, so shared navigation and search are generated under docs/.
• All pages are rendered with embedded resources so the checked-in HTML files remain self-contained.