MediStat Verifier
Compare results from Python and R
Review an existing analysis, reproduce it in R and compare the numerical results.
What to prepare
Source data, a Python analysis file and a study protocol if available
What you get
Analysis review · R reproduction · numerical comparison report
Requirements
Python and its dependencies; independent R reproduction also requires R and the relevant R packages
Python setup guide ↗Interpreting the results
This is an early-stage tool. Matching numbers do not establish that a statistical method is appropriate. Unsupported analyses require separate review.
Install in Claude Code
Enter these commands in the Code view of the Claude app or the Claude Code panel in VS Code. These are installation commands, not prompts for a regular Claude chat.
Run the first line, then the second. Restart Claude Code after installation to load the skill.
/plugin marketplace add JeonKH81/Medistat-verifier /plugin install medistat-verifier@medistat-verifier
These commands register the skill. Python, R and analysis packages must be prepared separately; check the requirements above and the original documentation.
Installation instructions on GitHub ↗Try this prompt after installation
Replace the bracketed fields with your own information, then paste the prompt into Claude Code.
[Goal] Check the computational reproducibility of an existing analysis. [Inputs] Data: [filename] Python analysis: [filename] Protocol: [filename or none] [Task] Use MediStat Verifier to review the methods and assumptions first, then reproduce supported analyses in R and report differences. [Checks] Present the verification plan and obtain approval before reproduction. Distinguish numerical agreement from methodological validity.
This page provides usage and installation guidance. Analyses and file processing take place in your installed Claude Code environment.