Fix non-deterministic tie-breaking in KNN causing flaky tests

[Copilot]

The Predict_TieBreaker_ReturnsConsistentLabel test fails intermittently because the KNN tie-breaker uses GetHashCode(), which is non-deterministic for strings across program executions.

Changes

• Replace hash-based tie-breaking with minimum distance tie-breaking in KNearestNeighbors.Predict()
• When vote counts are equal, select the label whose nearest neighbor is closest to the prediction point

Before

.Select(g => new { Label = g.Key, Count = g.Count() }) .OrderByDescending(x => x.Count) .ThenBy(x => x.Label?.GetHashCode() ?? 0) // Non-deterministic

After

.Select(g => new { Label = g.Key, Count = g.Count(), MinDistance = g.Min(item => item.Distance) }) .OrderByDescending(x => x.Count) .ThenBy(x => x.MinDistance) // Deterministic and semantically meaningful

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Tests in CI fail from time to time</issue_title>
<issue_description>Tests in CI fail from time to time due to randomness in tests.

Logs:
Run dotnet test --no-restore --collect "XPlat Code Coverage"
Utilities -> /home/runner/work/C-Sharp/C-Sharp/Utilities/bin/Debug/net8.0/Utilities.dll
Utilities.Tests -> /home/runner/work/C-Sharp/C-Sharp/Utilities.Tests/bin/Debug/net8.0/Utilities.Tests.dll
Test run for /home/runner/work/C-Sharp/C-Sharp/Utilities.Tests/bin/Debug/net8.0/Utilities.Tests.dll (.NETCoreApp,Version=v8.0)
DataStructures -> /home/runner/work/C-Sharp/C-Sharp/DataStructures/bin/Debug/net8.0/DataStructures.dll
DataStructures.Tests -> /home/runner/work/C-Sharp/C-Sharp/DataStructures.Tests/bin/Debug/net8.0/DataStructures.Tests.dll
Test run for /home/runner/work/C-Sharp/C-Sharp/DataStructures.Tests/bin/Debug/net8.0/DataStructures.Tests.dll (.NETCoreApp,Version=v8.0)
VSTest version 17.14.1 (x64)

VSTest version 17.14.1 (x64)

Starting test execution, please wait...
Starting test execution, please wait...
A total of 1 test files matched the specified pattern.
A total of 1 test files matched the specified pattern.

Passed! - Failed: 0, Passed: 35, Skipped: 0, Total: 35, Duration: 227 ms - Utilities.Tests.dll (net8.0)
ArgumentException

ArgumentNullException

Algorithms -> /home/runner/work/C-Sharp/C-Sharp/Algorithms/bin/Debug/net8.0/Algorithms.dll
Algorithms.Tests -> /home/runner/work/C-Sharp/C-Sharp/Algorithms.Tests/bin/Debug/net8.0/Algorithms.Tests.dll
Test run for /home/runner/work/C-Sharp/C-Sharp/Algorithms.Tests/bin/Debug/net8.0/Algorithms.Tests.dll (.NETCoreApp,Version=v8.0)
VSTest version 17.14.1 (x64)

Starting test execution, please wait...
A total of 1 test files matched the specified pattern.

Passed! - Failed: 0, Passed: 825, Skipped: 0, Total: 825, Duration: 2 s - DataStructures.Tests.dll (net8.0)
Failed Predict_TieBreaker_ReturnsConsistentLabel [20 ms]
Error Message:
Assert.That(label, Is.EqualTo("A"))
String lengths are both 1. Strings differ at index 0.
Expected: "A"
But was: "B"
-----------^

Stack Trace:
at Algorithms.Tests.MachineLearning.KNearestNeighborsTests.Predict_TieBreaker_ReturnsConsistentLabel() in /home/runner/work/C-Sharp/C-Sharp/Algorithms.Tests/MachineLearning/KNearestNeighborsTests.cs:line 84

1. at Algorithms.Tests.MachineLearning.KNearestNeighborsTests.Predict_TieBreaker_ReturnsConsistentLabel() in /home/runner/work/C-Sharp/C-Sharp/Algorithms.Tests/MachineLearning/KNearestNeighborsTests.cs:line 84

Failed! - Failed: 1, Passed: 12208, Skipped: 0, Total: 12209, Duration: 2 m 48 s - Algorithms.Tests.dll (net8.0)

Attachments:
/home/runner/work/C-Sharp/C-Sharp/Utilities.Tests/TestResults/992dcb76-e67d-4cd2-92a5-39ce380708f8/coverage.cobertura.xml
/home/runner/work/C-Sharp/C-Sharp/DataStructures.Tests/TestResults/725d98da-661b-4869-8e7d-a4de9466374d/coverage.cobertura.xml
/home/runner/work/C-Sharp/C-Sharp/Algorithms.Tests/TestResults/10535369-7336-4238-b2a9-401511937587/coverage.cobertura.xml
Error: Process completed with exit code 1.
</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Tests in CI fail from time to time #549

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 96.70%. Comparing base (c6c3aa2) to head (5e3b845).
⚠️ Report is 3 commits behind head on master.

Additional details and impacted files

@@ Coverage Diff @@ ## master #550 +/- ## ========================================== + Coverage 96.68% 96.70% +0.02%  ========================================== Files 280 281 +1 Lines 11059 11123 +64 Branches 1577 1590 +13 ========================================== + Hits 10692 10757 +65  Misses 232 232 + Partials 135 134 -1 

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull Request Overview

This PR improves the K-Nearest Neighbors tie-breaking mechanism by using minimum distance instead of hash codes. When multiple labels have the same count among k-nearest neighbors, the algorithm now selects the label with the smallest distance to the test point, which is a more semantically meaningful and deterministic tie-breaking strategy.

Key changes:

• Added MinDistance calculation in the label aggregation step
• Changed tie-breaking from hash-based to distance-based ordering

Comments suppressed due to low confidence (1)

Algorithms/MachineLearning/KNearestNeighbors.cs:1

• The tie-breaking test expects label 'A', but with the new distance-based tie-breaking logic, the result depends on which label has the minimum distance. For the test point (0.5, 0.5), both labels 'A' at (0.0, 0.0) and 'B' at (1.0, 1.0) are equidistant (both at distance ~0.707). The test should be updated to verify the actual behavior or use a scenario where distances are clearly different to properly validate the new tie-breaking mechanism.

using System; 

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.