Skip to content

[help-wanted-data] Calibrate disease prevalence to TÜİK national figures #23

Description

@ArioMoniri

Synthetic disease rates currently mirror the pristine-healthy source cohort, so HTN/DM/hyperlipidemia prevalence is well below Turkish population averages. Need: (a) TÜİK figures by age × sex for the 20 comorbidities, (b) a poststratification weighting scheme. Implementation lives between src/syntha/generator/copula.py (binary_p override) and a new src/syntha/calibration/tuik.py.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    help-wanted-dataNeeds a data scientist or statistician.

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions