[SoftwareX 2026] FastRerandomize: Rerandomization Using Accelerated Computing
Monte Carlo audit of naive, cluster-robust, and randomization-based experiment inference
Validation swaps
R package implementing Lee & Wooldridge (2025, 2026) rolling difference-in-differences estimator for panel data with staggered adoption, multiple estimators (RA/IPW/IPWRA/PSM), and exact small-sample inference.(Public Preview)(Disclaimer: CURRENTLY WIP)
EDI (Experimental Design and Inference): R & Python packages for fixed and sequential randomized experimental designs (blocking, optimal, matching-on-the-fly) with matched exact, asymptotic, and randomization inference for continuous, count, proportion, survival, and ordinal responses. C++ core. Not Electronic Data Interchange.
Exact randomization inference with uncertain record identities
Design-aware reanalysis of a household-randomized voter-turnout field experiment
Academic analysis of randomized-experiment data using treatment-effect estimation and randomization inference.
Replication package for spatial spillover analysis of Azerbaijan (RI placebo tests, spatial econometrics)
Out-of-sample evaluation of geopolitical risk as a carbon-price predictor, with a dimension-matched placebo design. Replication code for "Geopolitical Risk and Carbon Price Predictability: An Effect That Disappears."
Can public card data see the spending drop from ending pandemic unemployment benefits early? Staggered DiD across 48 states, 140 border-county pairs, randomization inference, power, and an audit folder.
Real-data evidence on intraday volatility periodicity across cryptocurrency assets, periods, and exchanges.
Replication code for "Can Fintech Build Trade Resilience Under Tariff Shocks?" — three-way FE PPML gravity on an East Asian trade panel, with N_eff (a treatment-concentration diagnostic), randomization inference, cluster jackknife, synthetic control, specification curve and MDE-based design requirements.
R tools for the Solomon four-group design: historical analyses, modern regression, randomization inference, maximum likelihood, SEM, power and sample-size planning, and Solomon-specific figures, each traced to its published source.
Do LLM-judge biases compose? A label-free, scale-aware test (SSIT) for position×length interaction in LLM-as-a-judge pipelines, with proofs, tie-coding and curvature diagnostics, 7-judge/6-family data, and the TMLR manuscript.
To associate your repository with the randomization-inference topic, visit your repo's landing page and select "manage topics."