Research Software
LookAgain-ML
A PyTorch-based library for extracting, comparing, selecting, and combining pretrained image representations, with group-aware validation, repeated-split stability, paired uncertainty, caching, and image quality-control workflows.
Implementations
lookagain-ml — Python GitHub only
Implementation: Vamuyan
Role: Creator and implementation author
Paper · Documentation · GitHub
Install
git clone https://github.com/vamusesay/LookAgain-ML.git
cd LookAgain-ML
python -m pip install .Associated paper
Nordstrom, Morgan, Ardyn Nordstrom, Matthew D. Webb, and Vamuyan Sesay. 2026. ‘Representation Risk in Pretrained Image Encoders.’ arXiv:2609.35470. Submitted.
How to cite
Please credit the software creator and cite ‘Representation Risk in Pretrained Image Encoders’ (arXiv:2609.35470) when the software contributes to published research.
summclust
Cluster-level leverage and influence diagnostics and cluster-jackknife variance estimation for linear regression models.
Implementations
summclust — Stata SSC
Implementation: Matthew D. Webb
Role: Original implementation
Paper · Documentation · GitHub · Video · SSC
Install
ssc install summclustAssociated paper
MacKinnon, James G., Morten Ørregaard Nielsen, and Matthew D. Webb. 2023. ‘Leverage, Influence, and the Jackknife in Clustered Regression Models: Reliable Inference Using summclust.’ Stata Journal 23(4): 942-982.
How to cite
Please cite the associated Stata Journal article when the method or software contributes to published research.
logitjack
Cluster-jackknife and wild-cluster-bootstrap inference for logistic regression models.
Implementations
logitjack — Stata SSC
Implementation: Matthew D. Webb
Role: Original implementation
Paper · Documentation · GitHub · Video · SSC
Install
ssc install logitjackAssociated paper
MacKinnon, James G., Morten Ørregaard Nielsen, and Matthew D. Webb. ‘Cluster-Robust Jackknife and Bootstrap Inference for Binary Response Models.’ Forthcoming, Econometric Reviews.
How to cite
Please cite the associated Econometric Reviews article; the SSC record also provides a software citation.
twowayjack
Jackknife variance estimation and inference for linear regression models with two-way clustering.
Implementations
twowayjack — Stata GitHub only
Implementation: Matthew D. Webb
Role: Original implementation
Paper · Documentation · GitHub
Install
net install twowayjack, from("https://raw.githubusercontent.com/mattdwebb/twowayjack/main/") replaceAssociated paper
MacKinnon, James G., Morten Ørregaard Nielsen, and Matthew D. Webb. 2024. ‘Jackknife Inference with Two-Way Clustering.’ arXiv:2406.08880.
How to cite
Please cite the associated paper when using the method or implementation.
mnwsvt
Score-variance tests for choosing the appropriate level of clustering in linear regression models.
Implementations
mnwsvt — Stata GitHub only
Implementation: Matthew D. Webb
Role: Original implementation
Paper · Documentation · GitHub
Install
net install mnwsvt, from("https://raw.githubusercontent.com/mattdwebb/mnwsvt/main/") replacemnwsvt — R GitHub only
Implementation: Tonghui Tian
Role: R implementation author
Paper · Documentation · GitHub
No package-manager installation is documented; use the upstream repository files directly.
Associated paper
MacKinnon, James G., Morten Ørregaard Nielsen, and Matthew D. Webb. 2023. ‘Testing for the Appropriate Level of Clustering in Linear Regression Models.’ Journal of Econometrics 235: 2027-2056.
How to cite
Please cite the associated Journal of Econometrics article.
CSDID cluster-jackknife inference
Cluster-jackknife (CV3) inference for the Callaway-Sant’Anna difference-in-differences estimator.
Implementations
csdidjack — Stata GitHub only
Implementation: Yunhan Liu
Role: Stata implementation author
Paper · Documentation · GitHub
Install
net install csdidjack, from("https://raw.githubusercontent.com/liu-yunhan/csdidjack/main/") replacedidjack — R GitHub only
Implementation: Yunhan Liu
Role: R implementation author
Paper · Documentation · GitHub
Install
remotes::install_github("liu-yunhan/didjack")fastcsdid — Stata GitHub
Implementation: Matthew D. Webb
Role: Stata implementation for stationary repeated cross sections, with analytical cluster and ordinary cluster-jackknife inference
Paper · Documentation · GitHub
Install
net install fastcsdid, from("https://raw.githubusercontent.com/mattdwebb/fastcsdid/main/stata") replacefastrcsdid — R GitHub
Implementation: Matthew D. Webb
Role: R implementation for stationary repeated cross sections, with analytical cluster and ordinary cluster-jackknife inference
Paper · Documentation · GitHub
Install
remotes::install_github("mattdwebb/fastcsdid", subdir = "R")fastcsdidpy — Python GitHub
Implementation: Matthew D. Webb
Role: Python implementation for stationary repeated cross sections, with analytical cluster and ordinary cluster-jackknife inference
Paper · Documentation · GitHub
Install
python -m pip install "git+https://github.com/mattdwebb/fastcsdid.git#subdirectory=python"Associated paper
Karim, Sunny R., Morten Ørregaard Nielsen, James G. MacKinnon, and Matthew D. Webb. 2026. ‘Improved Inference for CSDID Using the Cluster Jackknife.’ arXiv:2602.12043.
How to cite
Please cite the associated paper when using the method or software.
DID-INT
Intersection Difference-in-Differences estimation for settings in which covariate effects may vary across states, time, or both.
Implementations
DiDInt.jl — Julia Julia General Registry
Implementation: Eric Jamieson
Role: Julia implementation author
Paper · Documentation · GitHub · Video · Julia General Registry
Install
using Pkg
Pkg.add("DiDInt")Associated paper
Karim, Sunny, and Matthew D. Webb. 2024. ‘Good Controls Gone Bad: Difference-in-Differences with Covariates.’ arXiv:2412.14447.
How to cite
Please cite the associated paper when using the method or software.
UN-DID
Difference-in-differences for data held in separate silos that cannot be pooled, with implementations for four programming languages.
Implementations
undid — Stata SSC
Implementation: Eric Jamieson
Role: Stata implementation author
Paper · Documentation · GitHub · SSC
Install
ssc install undidundidR — R CRAN
Implementation: Eric Jamieson
Role: R implementation author
Paper · Documentation · GitHub · CRAN
Install
install.packages("undidR")Undid.jl — Julia GitHub only
Implementation: Eric Jamieson
Role: Julia implementation author
Paper · Documentation · GitHub
Install
using Pkg
Pkg.add(url="https://github.com/ebjamieson97/Undid.jl")undidPyjl — Python PyPI
Implementation: Eric Jamieson
Role: Python wrapper author
Paper · Documentation · GitHub · PyPI
Install
pip install undidPyjlAssociated paper
Karim, Sunny, Nichole Austin, Erin Strumpf, and Matthew D. Webb. 2025. ‘Difference-in-Differences with Unpoolable Data.’ arXiv:2403.15910.
How to cite
Please cite the associated paper when using the method or software.
boottest
Fast wild-bootstrap inference for linear regression models in Stata.
Implementations
boottest — Stata SSC
Implementation: David Roodman
Role: Original developer and Stata implementation author
Paper · Documentation · GitHub · Video · SSC
Install
ssc install boottestAssociated paper
Roodman, David, James G. MacKinnon, Morten Ørregaard Nielsen, and Matthew D. Webb. 2019. ‘Fast and Wild: Bootstrap Inference in Stata Using boottest.’ Stata Journal 19(1): 4–60.
How to cite
Please cite the associated Stata Journal article when using boottest.