feat(scoring): apply sqrt damping to density calculation#227
Merged
Conversation
Same rule triggered N times now contributes |score| × sqrt(N) instead of |score| × N. This prevents repeated issues (e.g., raw-value ×79) from destroying category scores while preserving sensitivity for rare issues where sqrt ≈ linear. Before: Community design system files score D/F (45-57%) After: Same files score B/C+ (72-78%) https://claude.ai/code/session_01H2EPyD3PhqkQFrk5BB4M4g
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughDensity scoring changed from summing per-issue absolute scores to per-rule square-root damping: issues are grouped by Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 4✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
This was referenced Mar 31, 2026
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
src/core/engine/scoring.test.ts (1)
119-135: 🧹 Nitpick | 🔵 TrivialConsider adding an explicit test for sqrt damping with multiple same-rule issues.
The test verifies relative ordering (many > few issues decreases density score), but doesn't assert the exact
weightedIssueCountvalue that demonstrates sqrt damping. Adding an assertion would strengthen confidence in the formula:// 5 issues of rule with score -5: weightedIssueCount = |5| × sqrt(5) ≈ 11.18 expect(many.byCategory["pixel-critical"].weightedIssueCount).toBeCloseTo(5 * Math.sqrt(5), 1);🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/core/engine/scoring.test.ts` around lines 119 - 135, Add an explicit assertion to the test that verifies the sqrt damping calculation for multiple identical-rule issues: after computing `many = calculateScores(makeResult([...], 100))`, assert that `many.byCategory["pixel-critical"].weightedIssueCount` is approximately equal to 5 * Math.sqrt(5) (use toBeCloseTo with an appropriate precision), in addition to the existing densityScore comparison; this uses the existing helpers `calculateScores`, `makeResult`, and `makeIssue`.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Outside diff comments:
In `@src/core/engine/scoring.test.ts`:
- Around line 119-135: Add an explicit assertion to the test that verifies the
sqrt damping calculation for multiple identical-rule issues: after computing
`many = calculateScores(makeResult([...], 100))`, assert that
`many.byCategory["pixel-critical"].weightedIssueCount` is approximately equal to
5 * Math.sqrt(5) (use toBeCloseTo with an appropriate precision), in addition to
the existing densityScore comparison; this uses the existing helpers
`calculateScores`, `makeResult`, and `makeIssue`.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: dbface2f-dae1-49c0-8147-4cb8f38e831b
📒 Files selected for processing (2)
src/core/engine/scoring.test.tssrc/core/engine/scoring.ts
Add toBeCloseTo assertion verifying weightedIssueCount = |score| × sqrt(count) for repeated same-rule issues. https://claude.ai/code/session_01H2EPyD3PhqkQFrk5BB4M4g
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
src/core/engine/scoring.test.ts (1)
119-137: 🧹 Nitpick | 🔵 TrivialOptional: Consider adding multi-rule sqrt damping test.
The current test coverage is solid and validates the core sqrt damping behavior. For comprehensive coverage, you might consider adding a test that explicitly verifies sqrt damping across multiple rules in the same category, e.g.:
it("applies sqrt damping independently per rule", () => { const issues = [ // 4 issues of rule A (score -5) ...Array(4).fill(null).map(() => makeIssue({ ruleId: "no-auto-layout", category: "pixel-critical", severity: "blocking", score: -5 })), // 9 issues of rule B (score -3) ...Array(9).fill(null).map(() => makeIssue({ ruleId: "non-layout-container", category: "pixel-critical", severity: "risk", score: -3 })), ]; const scores = calculateScores(makeResult(issues, 100)); // Expected: 5×sqrt(4) + 3×sqrt(9) = 5×2 + 3×3 = 10 + 9 = 19 expect(scores.byCategory["pixel-critical"].weightedIssueCount).toBeCloseTo(19, 1); });However, this is truly optional—the existing tests adequately cover the sqrt damping feature.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/core/engine/scoring.test.ts` around lines 119 - 137, Add an optional unit test that verifies sqrt damping is applied per-rule within a category by creating multiple issues across different ruleIds and asserting the category weightedIssueCount equals the sum of each rule's (abs(score) × sqrt(count))—for example, create 4 issues with score -5 for ruleId "no-auto-layout" and 9 issues with score -3 for ruleId "non-layout-container", call calculateScores(makeResult(issues, 100)) and assert scores.byCategory["pixel-critical"].weightedIssueCount isCloseTo( (5 * Math.sqrt(4)) + (3 * Math.sqrt(9)), 1 ); use existing helpers makeIssue, makeResult and calculateScores and name the test something like "applies sqrt damping independently per rule".
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/core/engine/scoring.test.ts`:
- Around line 89-102: Update the inline comment near the test case that sets
heavyIssue.calculatedScore to -15 to explicitly state that the entire
calculatedScore field is ignored for density calculations (not just the depth
component); modify the comment in the test around heavyIssue/calculatedScore in
scoring.test.ts to read something like "calculatedScore is ignored for density —
density uses the base score (score = -10), so |−10| × sqrt(1) = 10" to make
intent clear when reading the calculateScores/byCategory assertions.
---
Outside diff comments:
In `@src/core/engine/scoring.test.ts`:
- Around line 119-137: Add an optional unit test that verifies sqrt damping is
applied per-rule within a category by creating multiple issues across different
ruleIds and asserting the category weightedIssueCount equals the sum of each
rule's (abs(score) × sqrt(count))—for example, create 4 issues with score -5 for
ruleId "no-auto-layout" and 9 issues with score -3 for ruleId
"non-layout-container", call calculateScores(makeResult(issues, 100)) and assert
scores.byCategory["pixel-critical"].weightedIssueCount isCloseTo( (5 *
Math.sqrt(4)) + (3 * Math.sqrt(9)), 1 ); use existing helpers makeIssue,
makeResult and calculateScores and name the test something like "applies sqrt
damping independently per rule".
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: 8497d683-3d98-4b55-be79-9a575d6c6e73
📒 Files selected for processing (1)
src/core/engine/scoring.test.ts
- Clarify comment: calculatedScore is fully ignored for density - Add multi-rule test verifying per-rule independent damping https://claude.ai/code/session_01H2EPyD3PhqkQFrk5BB4M4g
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
|score| × sqrt(N)instead of|score| × NBefore → After (5 real fixtures)
Why
raw-value-4 × 79건 = -316 weight against 273 nodes → density destroyed.With sqrt: -4 × √79 ≈ -36 → reasonable penalty for a systemic issue.
B→A remains hard (low issue counts, sqrt ≈ linear) — this is intentional. A requires genuinely well-structured designs.
Test plan
pnpm test:run)pnpm lintpassesCloses #226
https://claude.ai/code/session_01H2EPyD3PhqkQFrk5BB4M4g
Summary by CodeRabbit