TaskFactorStudyScenario
Controlled_Studies scenarios · Implementation: mlsysim/engine/scenario_definitions/controlled_studies.py.
Construct this scenario with keyword inputs. The fields below show the types and defaults declared by this class.
A bounded controlled study with authored task cards.
Notes
Equation: quality passes iff supplied quality is at least quality_floor_pct. Assumptions: Cards describe one fixed host and carry independent quality inputs. Excludes: Learned quality prediction, selection overhead, and device calibration.
References
Derivation: Compare a separately supplied quality observation with the fixed acceptance floor; missing observations cannot establish acceptance.
Examples
>>> from mlsysim import Scenarios
>>> Scenarios.ControlledTaskFactors.cards["baseline"].samples > 0
TrueAttributes
| Name | Description |
|---|---|
| capability_options | |
| cards | |
| provenance | |
| quality_floor_pct | |
| shift_options | |
| version |
version
Type: str.
Declared default: required.
quality_floor_pct
Type: float.
Declared default: required.
shift_options
Type: tuple[int, ...].
Declared default: required.
capability_options
Type: tuple[float, ...].
Declared default: required.
cards
Type: Dict[str, TaskFactorCard].
Declared default: required.
provenance
Type: Provenance.
Declared default: required.