Skip to content

Constants

File: src/constants.py

EndpointType

An enum that defines the three supported outcome endpoint types. Used throughout the pipeline to select the correct model class and validation rules.

from src.constants import EndpointType
Member Value Meaning
EndpointType.LOGICAL "logical" Binary classification (0/1 outcome)
EndpointType.INTEGER "integer" Count regression (non-negative integer outcome)
EndpointType.SURVIVAL "survival" Time-to-event (duration + event indicator)

Usage

from src.constants import EndpointType

# Construct an outcome config
outcome = {
    "endpointType": EndpointType.LOGICAL,
    "columnsToUse": ["ctn0094_relapse_event"]
}

# Compare
if outcome["endpointType"] == EndpointType.LOGICAL:
    print("Binary classification task")

# Instantiate from string
endpoint = EndpointType("logical")  # → EndpointType.LOGICAL

Note

Passing an invalid string to EndpointType() raises a ValueError. All pipeline entry points validate the endpoint type before any data processing begins.