Metric cards and the MCDA pipeline
A metric card declares fields for identity, kind, inputs, output, semantics, comparability, implementations, examples, and provenance. See measurement theory and normalization and scales for the scale types and the normalization strategies.
Fields read
polarity:normalizeinvertslower_is_bettercolumns and rescales each column to [0, 1], higher is better.range_lower,range_upper: when both are declared,run_from_registrypasses them tonormalize, and min-max and baseline-relative use them instead of the observed extrema. A value outside the declared range raises an error.comparability.recommended_normalization: the strategy for the column, by defaultmin_max. The options aremin_max,log_min_max,rank,zscore,baseline_relativeandtarget_relative.semantics.score_of_random_baseline: the chance-level value, used bybaseline_relative.semantics.target: the ideal value, used bytarget_relative.scale_type:validate_for_aggregationrefusesnominalandordinalcolumns for every aggregation.allowed_transformations:validate_for_aggregationchecks that the card allows the transform of the strategy. Min-max, baseline-relative and target-relative needaffineormin_max;log_min_maxneedslog;rankneedsrank;zscoreneedsz_scoreoraffine.comparability.recommended_aggregation_across_datasets:aggregate_across_datasetsuses it to reduce a tool by dataset matrix to one value per tool. Runtime and peak memory declaregeometric_mean(Smith 1988, 10.1145/63039.63043); bounded metrics declarearithmetic_mean.
For a min-max column run_from_registry warns when a declared bound is missing or the column is heavy-tailed. The warnings are in the Result and the run continues.
Fields declared but not enforced
meaningful_zeroandmonotonic: no reader.uncertainty_model: uncertainty is not propagated through aggregation.comparability.comparable_withinandaggregation_rules: notes for readers.