statistical-analysis

Automate regression and survival analysis by loading a prepared dataset alongside a research_questions.json file. The workflow identifies and executes the correct statistical analyses, generating structured analysis_results.json and accompanying reports for reproducible research outputs.

Causal inference and biostatistics across longitudinal data require inferential methods like regression and survival analysis. The workflow applies these methods automatically to cross-sectional and longitudinal designs, producing standardized reports and sensitivity checks.

Yes, you can use statsmodels and lifelines with pandas for reproducible statistical analysis. The workflow requires a Python environment with these libraries to drive descriptive statistics, causal inference, and survival analyses from prepared datasets into structured outputs.

Generate Table 1 descriptive statistics and sensitivity checks by running the end-to-end statistical pipeline. It orchestrates data preparation, analytic dataset creation, and analysis plan execution to produce publication-quality reporting and auditability.

No, xgboost and scikit-learn are optional machine learning libraries for the workflow. Core causal inference and biostatistics workflows run on required dependencies like numpy, pandas, scipy, statsmodels, and lifelines.

Statistical analysis pipelines fail to produce analysis_results.json outputs if the Python environment lacks required dependencies like statsmodels or lifelines, or if the prepared dataset and research_questions.json file are not loaded correctly into the workflow.