spb/tools-spboucher-ai

1{2  "slug": "fixest",3  "name": "fixest",4  "aliases": [],5  "vendor": { "name": "Laurent Bergé", "slug": "laurent-berge", "website": "https://lrberge.github.io/fixest/" },6  "homepage_url": "https://lrberge.github.io/fixest/",7  "docs_url": "https://lrberge.github.io/fixest/",8  "repo_url": "https://github.com/lrberge/fixest",9  "license_url": "https://cran.r-project.org/package=fixest",10  "product_type": "library",11  "product_status": "active",12  "summary": "R package for fast estimation of OLS, GLM, Poisson, IV and difference-in-differences models with many fixed effects, clustered errors, multiple estimations and publication tables",13  "description": "fixest is an R package by Laurent Bergé for fast econometric estimation with high-dimensional fixed effects, released under the GPL-3 and available on CRAN. A single formula syntax separates regressors, fixed effects and instruments with pipes and covers ordinary least squares (feols), generalized linear models and Poisson pseudo-maximum likelihood (feglm, fepois), negative binomial, maximum-likelihood models and two-stage least squares, with fixed effects absorbed by an efficient demeaning algorithm that handles millions of observations and several fixed-effect dimensions on a laptop.\n\nThe package is built for applied-microeconomics workflows: six families of variance-covariance estimators including multi-way clustering and Driscoll-Kraay, one-call multiple estimations over sets of dependent variables, regressors, fixed effects and sample splits, lead/lag/difference operators for panels, event-study estimation with interaction syntax and the Sun-Abraham estimator for staggered adoption, coefficient and event-study plots, and etable, which prints regression tables to the console or exports publication-ready LaTeX with full styling control. Version 0.14.2 is current at the time of writing.\n\nfixest has become the default fixed-effects engine in R, displacing lfe, and its syntax has been ported to Python as pyfixest. It is single-machine and in-memory, offers no nonlinear panel estimators beyond GLM families, and its many options and compact syntax have a learning curve; users still reach for other packages for synthetic control or modern heterogeneous-treatment DID estimators.",14  "capabilities": ["feols, feglm, fepois and fenegbin with absorbed fixed effects", "Instrumental-variable (2SLS) estimation within the same syntax", "Multi-way clustered, heteroskedasticity-robust and Driscoll-Kraay standard errors", "Multiple estimations in one call over variables, fixed effects and samples", "Event-study and Sun-Abraham staggered DID estimation", "Panel lead, lag and difference operators", "etable for console and LaTeX regression tables", "Coefficient and event-study plots"],15  "workflows": ["Two-way fixed-effects regressions with clustered errors for a paper", "Staggered difference-in-differences with event-study plots", "Gravity models with PPML and exporter-importer-year fixed effects", "Robustness tables across many specifications exported to LaTeX"],16  "advantages": ["Fast on large panels with many fixed effects", "Compact formula syntax for complex designs", "Publication tables without extra packages", "GPL-3 open source on CRAN"],17  "limitations": ["R only (Python port is a separate project)", "Single-machine, in-memory", "Limited nonlinear models beyond GLM families", "Dense option set has a learning curve"],18  "prerequisites": ["R with CRAN access"],19  "audiences": ["Applied micro and trade economists", "Graduate students writing empirical papers", "Researchers replacing Stata reghdfe workflows in R"],20  "alternatives": ["statsmodels", "linearmodels", "stata", "r"],21  "comparison_notes": "linearmodels and statsmodels in Python offer fixed-effects and IV estimation but without fixest's speed on many-level effects or its multiple-estimation and table tooling; Stata's reghdfe is the closest equivalent in a commercial package. pyfixest (not recorded here) ports the syntax to Python.",22  "categories": [{ "slug": "econometrics-causal-inference", "primary": true }],23  "tags": ["r", "econometrics", "causal-inference", "panel-data", "open-source"],24  "use_cases": ["estimate-causal-effects", "write-papers"],25  "operating_systems": ["macos", "linux", "windows"],26  "execution_modes": ["local"],27  "interfaces": ["library"],28  "deployment_models": ["embedded"],29  "integrations": ["R", "data.table", "ggplot2", "modelsummary", "RStudio", "Quarto"],30  "data_formats": ["R data frames", "data.table", "LaTeX tables"],31  "languages": ["R", "C++"],32  "source_model": "open-source",33  "license_spdx": "GPL-3.0-only",34  "pricing_model": ["free", "open-source"],35  "self_hostable": true,36  "sources": [37    { "url": "https://lrberge.github.io/fixest/", "type": "official_site", "title": "Fast Fixed-Effects Estimations • fixest", "supports": ["name", "summary", "capabilities", "vendor", "license_spdx"], "checked_at": "2026-11-02" },38    { "url": "https://github.com/lrberge/fixest", "type": "repository", "title": "lrberge/fixest", "supports": ["repo_url", "license_spdx", "languages"], "checked_at": "2026-11-02", "note": "DESCRIPTION file declares License: GPL-3" },39    { "url": "https://cran.r-project.org/package=fixest", "type": "package_registry", "title": "CRAN - Package fixest", "supports": ["license_spdx", "product_status"], "checked_at": "2026-11-02" }40  ],41  "confidence": "high",42  "discovery_method": "editorial-research",43  "depth": "deep",44  "editorial_status": "reviewed",45  "visibility": "public",46  "featured": false,47  "author": "batch-E"48}49