ProgramMetricIssueReason

Enums · AiDotNet.Enums

Explains why one metric did not contribute to a scalar quality in the way a caller might expect.

For Beginners

When several measurements are combined into a single score, some of them may be unusable - a message instead of a number, a yes/no flag, or a value that is missing. Rather than quietly ignoring those, the aggregation returns a list of them with one of these reasons attached, so you can see at a glance why your score is not what you expected.

How It Works

The reference OpenEvolve scalarization silently discards every value it cannot use: a string metric, a boolean flag, a NaN, or a metric excluded because it is an archive feature dimension all vanish without trace, and an evaluator whose numbers never reach the fitness score looks identical to one whose numbers did. Each member of this enumeration names one of those discards so the aggregation can report it instead.

Fields

Field Summary
ProgramMetricIssueReason BooleanFlag A boolean flag was skipped because a flag is not a score.
ProgramMetricIssueReason ExcludedFeatureDimension A metric was excluded from the aggregation because it is an archive feature dimension.
ProgramMetricIssueReason MissingMetric A metric named by the configuration was absent from the reported dictionary.
ProgramMetricIssueReason NoNumericValues No usable numeric metric was found at all, so the aggregation fell back to zero.
ProgramMetricIssueReason NoWeightDeclared A numeric metric was reported but the weighted strategy declares no weight for it.
ProgramMetricIssueReason NonNumericText A text metric was skipped because it could not be read as a finite number.
ProgramMetricIssueReason NotANumber A numeric metric was skipped because its value was not a number.
ProgramMetricIssueReason NotFinite A numeric metric was infinite, so the aggregated value cannot be finite.
int value__