Skip to content

API Reference

This section contains the complete API documentation for CLIFpy, automatically generated from the source code docstrings.

Core Components

ClifOrchestrator

The main orchestration class for managing multiple CLIF tables with consistent configuration.

BaseTable

The base class that all CLIF table implementations inherit from, providing common functionality for data loading, validation, and reporting.

Table Classes

Tables Overview

Complete API documentation for all CLIF table implementations:

  • Patient - Patient demographics and identification
  • Adt - Admission, discharge, and transfer events
  • Hospitalization - Hospital stay information
  • Labs - Laboratory test results
  • Vitals - Vital signs measurements
  • RespiratorySupport - Ventilation and oxygen therapy
  • MedicationAdminContinuous - Continuous medication infusions
  • PatientAssessments - Clinical assessment scores
  • Position - Patient positioning data

DQA (Validation)

DQA API Reference

Comprehensive Data Quality Assessment framework with 30+ checks across three pillars:

  • Conformance — Schema validation, data types, categorical values, datetime formats
  • Completeness — Missingness analysis, conditional requirements, relational integrity
  • Plausibility — Temporal ordering, numeric ranges, duplicate detection, cross-table consistency
  • Orchestration — run_full_dqa, category-level runners, cache-based cross-table pipeline
  • Report Generation — PDF and text report output from DQA results

Utilities

Utility Functions

Helper functions for data processing and specialized operations:

  • stitch_encounters — Link related hospitalizations within time windows
  • process_resp_support_waterfall — Respiratory support waterfall algorithm
  • compute_ase — CDC Adult Sepsis Event surveillance calculation
  • io — Data loading and sample creation utilities
  • config — Configuration management functions
  • outlier_handler — Outlier detection and handling
  • wide_dataset — Wide dataset creation utilities

Usage Example

from clifpy.clif_orchestrator import ClifOrchestrator
from clifpy.tables import Patient, Labs, Vitals

# Using the orchestrator
orchestrator = ClifOrchestrator(
    data_directory='/path/to/data',
    filetype='parquet',
    timezone='US/Central'
)
orchestrator.initialize(tables=['patient', 'labs', 'vitals'])

# Using individual tables
patient = Patient.from_file('/path/to/data', 'parquet')
patient.validate()