Data Analysis Generate rigorous statistical analysis code with multi-round review. Input - — Data source (CSV, JSON, pickle, or experiment logs) - — Research goal or hypothesis to test References - 4-round code review prompts: Scripts Statistical summary and comparison Detects data types, recommends tests, runs comparisons, outputs effect sizes and significance stars. Requires numpy, scipy. Format p-values Formats p-values with stars, LaTeX notation, or plain text. Stdlib-only. Workflow Step 1: Generate Analysis Code Structure the code with these sections: 1. — pandas, numpy, scipy, statsmode…