Skip to content

setup.md

Latest commit

 

History

History
45 lines (30 loc) · 1.29 KB

setup.md

File metadata and controls

45 lines (30 loc) · 1.29 KB
Raw

Setup & Build

1. Clone the repository with submodules

AlgoGauge depends on two git submodules:

  • external/trade-ngin – the trading engine library being benchmarked
  • tools/FlameGraph – Brendan Gregg's flamegraph generation scripts

Clone both at once:

git clone --recurse-submodules https://github.com/AlgoGators/algogauge.git
cd algogauge

If you already cloned without --recurse-submodules, initialize them now:

git submodule update --init --recursive

2. Install Python dependencies

uv sync

This creates a virtual environment and installs all packages listed in pyproject.toml (Dash, Pandas, Plotly, etc.) using the exact versions pinned in uv.lock.

3. Build the C++ benchmark binaries

cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build --parallel

The compiled binaries are placed in build/benchmarks/:

Binary Strategy
build/benchmarks/base_strategy_benchmarks trade_ngin::BaseStrategy
build/benchmarks/trend_following_benchmarks trade_ngin::TrendFollowingStrategy

Tip: Pass -DCMAKE_BUILD_TYPE=RelWithDebInfo instead of Release to keep debug symbols while still benefiting from optimizations. This produces richer flamegraphs without changing benchmark results.