BiasAnalyzerCohortDefinitionBuilder

A Python-based package for generating Cohort definition files compatible with BiasAnalyzer.
It allows researchers to construct complex cohort logic (e.g., temporal relationships, event sequences, logical filters) directly from Python.

---

Features

• Event primitives
  • ConditionOccurrence (Dx)
  • DrugExposure (Rx)
  • Measurement (Labs)
  • VisitOccurrence (encounters)
  • DateEvent (Fixed calendar dates)
• Attribute primitives
  • Demographics (gender, birth year, race, ethnicity)
• Logical operators
  • AND, OR, NOT — for Boolean logic
  • BEFORE — for temporal relationships
• Automatic YAML serialization
  Generate ready-to-use .yaml cohort definition files directly from Python objects.
• Flexible schema handling
  Fully aligned with BiasAnalyzer’s cohort schema — no structural modifications required.

---

Project Structure

CohortDefinition/
├── CohortDefinition/
│   ├── __init__.py
│   ├── builder.py              # Core Cohort builder & CohortCriteria class
│   ├── events.py               # Event primitives (Dx, Encounters, etc.)
│   └── logic.py                # Logical & temporal operators
├── examples/
│   ├── build_example1.py
│   ├── build_example2.py
│   ├── build_example3_no_demo.py
│   ├── build_example4_no_temporal.py
│   └── build_example5_not.py
├── setup.py
├── README.md
└── LICENSE