無料GitHub で公開

performance-testing

Benchmark indicator performance with BenchmarkDotNet. Use for Series/Buffer/Stream benchmarks, regression detection, and optimization patterns. Target 1.5x Series for StreamHub, 1.2x for BufferList.

foryourhealth111-pixel/Vibe-Skills

インストール方法を見る

含まれるファイル（2）

SKILL.md3.1 KB
references/benchmark-patterns.md3.6 KB

SKILL.md（原文）

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Performance testing

Running benchmarks

cd tools/performance

# Run all benchmarks (~15-20 minutes)
dotnet run -c Release

# Run specific category
dotnet run -c Release --filter *StreamIndicators*
dotnet run -c Release --filter *BufferIndicators*
dotnet run -c Release --filter *SeriesIndicators*

# Run specific indicator
dotnet run -c Release --filter *.EmaHub

Adding benchmarks

Series pattern

[Benchmark]
public void ToMyIndicator() => quotes.ToMyIndicator(14);

Stream pattern

[Benchmark]
public object MyIndicatorHub() => quoteHub.ToMyIndicatorHub(14).Results;

Buffer pattern

[Benchmark]
public MyIndicatorList MyIndicatorList() => new(14) { quotes };

Style comparison

[Benchmark]
public IReadOnlyList<MyResult> MyIndicatorSeries() => quotes.ToMyIndicator(14);

[Benchmark]
public IReadOnlyList<MyResult> MyIndicatorBuffer() => quotes.ToMyIndicatorList(14);

[Benchmark]
public IReadOnlyList<MyResult> MyIndicatorStream() => quoteHub.ToMyIndicator(14).Results;

Performance targets

Note: These are optimization goals for future v3.1+ effort. Current implementations vary—see PERFORMANCE_ANALYSIS.md for actual measured performance. Some indicator families (e.g., EMA) have inherent framework overhead due to simple operation costs.

Style	Target vs Series	Use Case
Series	Baseline	Batch processing
BufferList	≤ 1.2x	Incremental data
StreamHub	≤ 1.5x	Real-time feeds

Expected execution times (502 periods)

Note: These are optimization targets. Actual execution times vary by indicator complexity and current implementation.

Complexity	Time	Examples
Fast	< 30μs	SMA, EMA, WMA, RSI
Medium	30-60μs	MACD, Bollinger Bands, ATR
Complex	60-100μs	HMA, ADX, Stochastic
Advanced	100-200μs+	Ichimoku, Hurst

Regression detection

# Auto-detect baseline and results
pwsh detect-regressions.ps1

# Custom threshold (default 10%)
pwsh detect-regressions.ps1 -ThresholdPercent 15

Exit codes:

0 - No regressions
1 - Regressions found

Creating baselines

cp BenchmarkDotNet.Artifacts/results/Performance.*-report-full.json \
   baselines/baseline-v3.0.0.json

Required optimization patterns

Minimize allocations in hot paths
Avoid LINQ in performance-critical loops
Use Span<T> for zero-copy operations
Cache calculations when possible
Test with realistic data sizes (502 periods)

Prohibited patterns

Excessive LINQ in hot paths
Boxing/unboxing of value types
Unnecessary string allocations
Redundant calculations in loops
Poor cache locality

See references/benchmark-patterns.md for detailed patterns.

Last updated: December 31, 2025

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

ログインしてレビューを書く

同じリポジトリのスキル

概要と使いどころ

adaptyv

無料

Cloud laboratory platform for automated protein testing and validation. Use when designing proteins and needing experimental validation including binding assays, expression testing, thermostability measurements, enzyme activity assays, or protein sequence optimization. Also use for submitting experiments via API, tracking experiment status, downloading results, optimizing protein sequences for better expression using computational tools (NetSolP, SoluProt, SolubleMPNN, ESM), or managing protein design workflows with wet-lab validation.

日本語の概要は準備中です。原文の説明を表示しています。