Precision and recall by outcome-to-label pairs from labeled events
Hide low-support pairs where both predicted and actual counts are below threshold.
{{ polling ? 'Generating snapshot report...' : 'Loading rule quality report...' }}
Labeled Events
{{ totalLabeledEvents }}
Rules Analyzed
{{ uniqueRulesCount() }}
Outcome/Label Pairs
{{ pairMetrics.length }}
Highest average F1 score
| Rule | Avg F1 | Best Pair |
|---|---|---|
| {{ rule.rid }} | {{ formatPercent(rule.averageF1) }} | {{ rule.bestPair || 'N/A' }} |
| No ranked rules for this support threshold. | ||
Lowest average F1 score
| Rule | Avg F1 | Worst Pair |
|---|---|---|
| {{ rule.rid }} | {{ formatPercent(rule.averageF1) }} | {{ rule.worstPair || 'N/A' }} |
| No ranked rules for this support threshold. | ||
Each row evaluates one mapping: outcome prediction vs ground-truth label.
| Rule | Outcome | Label | Precision | Recall | F1 | TP | FP | FN | Predicted | Actual |
|---|---|---|---|---|---|---|---|---|---|---|
| {{ metric.rid }} | {{ metric.outcome }} | {{ metric.label }} | {{ formatPercent(metric.precision) }} | {{ formatPercent(metric.recall) }} | {{ formatPercent(metric.f1) }} | {{ metric.truePositive }} | {{ metric.falsePositive }} | {{ metric.falseNegative }} | {{ metric.predictedPositives }} | {{ metric.actualPositives }} |