Returns a position profile across archetypes — NOT a pass/fail gate.
See concept note Stage 0.75 for the double-dipping caveat and the
bootstrap vs. SBC split.
Usage
assess_fit(strategy, data, archetype, ...)
Arguments
- strategy
a FitAssessor implementation
- data
StateTransitionData
- archetype
character archetype name to assess against
- ...
forwarded to the implementation
Details
Return value (in StageResult@value):
- null_position
named numeric: percentile in each archetype's null
- assignment_probs
named numeric: calibrated posterior probabilities
- uncertainty
numeric: spread of assignment probabilities