Workflow

  1. Upload or select a sample geo-level panel dataset.
  2. Map location, date, and outcome columns.
  3. Run GeoDataRead-style preprocessing.
  4. Configure budget, CPIC, effect size, holdout, treatment periods, and location filters.
  5. Run market selection and choose a candidate treatment group.
  6. Run final GeoLift treatment analysis and export results.

Method Note

Deterministic quantities are designed to mirror R GeoLift closely. The Python package uses its documented conformal inference path for p-values and confidence intervals, so those values may differ from R GeoLift.