lean-engine（精益发动机）

LEAN Engine — QuantConnect Algorithmic Trading

Prerequisites & Setup

Required Environment Variables

All three must be set before using this skill. Add to your shell profile:

export LEAN_ROOT="$HOME/lean"
export DOTNET_ROOT="$HOME/.dotnet"
export PATH="$PATH:$DOTNET_ROOT"
export PYTHONNET_PYDLL="$LEAN_ROOT/.libs/libpython3.11.so.1.0"

Note: LEAN bundles its own Python shared library in $LEAN_ROOT/.libs/. If you built LEAN from source, the library should be there after dotnet build. If not, install libpython3.11-dev and point PYTHONNET_PYDLL to your system's libpython3.11.so.

First-Time Setup

1. Install .NET 8 SDK:

   # Linux/macOS
   wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh
   chmod +x dotnet-install.sh
   ./dotnet-install.sh --channel 8.0
   export DOTNET_ROOT="$HOME/.dotnet"
   export PATH="$PATH:$DOTNET_ROOT"

1. Clone and build LEAN:

   git clone https://github.com/QuantConnect/Lean.git "$LEAN_ROOT"
   cd "$LEAN_ROOT"
   dotnet build QuantConnect.Lean.sln -c Debug

1. Download initial market data:

   pip install yfinance pandas
   python3 {baseDir}/scripts/download_us_universe.py --symbols sp500 --start 2020-01-01 --data-dir "$LEAN_ROOT/Data"

1. Verify setup:

   ls "$LEAN_ROOT/Data/equity/usa/daily/"  # Should list .zip files
   ls "$LEAN_ROOT/Launcher/bin/Debug/"      # Should contain QuantConnect.Lean.Launcher.dll

Environment

• LEAN source: $LEAN_ROOT/
• Launcher (pre-built): $LEAN_ROOT/Launcher/bin/Debug/
• Config: $LEAN_ROOT/Launcher/config.json
• Python algos: $LEAN_ROOT/Algorithm.Python/
• Market data: $LEAN_ROOT/Data/
• dotnet: $DOTNET_ROOT/dotnet (add to PATH: export PATH="$PATH:$DOTNET_ROOT")

Quick Reference

Run a Backtest

1. Place algorithm in $LEAN_ROOT/Algorithm.Python/YourAlgo.py
2. Edit config to point to it:

   # Update config.json — set these fields:
   # "algorithm-type-name": "YourClassName"
   # "algorithm-language": "Python"
   # "algorithm-location": "../../../Algorithm.Python/YourAlgo.py"

1. Run:

   export PATH="$PATH:$DOTNET_ROOT"
   cd "$LEAN_ROOT/Launcher/bin/Debug"
   dotnet QuantConnect.Lean.Launcher.dll

1. Results appear in stdout + $LEAN_ROOT/Results/

Or use the helper script:

bash {baseDir}/scripts/run_backtest.sh YourClassName YourAlgo.py

Config Editing

Edit $LEAN_ROOT/Launcher/config.json with these key fields:

For IB live trading, set environment to "live-interactive" and configure the ib-* fields (account, username, password, host, port, trading-mode).

Data Management

Check available data:

ls "$LEAN_ROOT/Data/equity/usa/daily/"

Data format: ZIP files containing CSV. Each line: YYYYMMDD HH:MM,Open*10000,High*10000,Low*10000,Close*10000,Volume

Prices are stored as integers (multiply by 10000). LEAN handles conversion internally.

Download more data:

python3 {baseDir}/scripts/download_us_universe.py --symbols sp500 --data-dir "$LEAN_ROOT/Data"

See {baseDir}/references/data-download.md for additional methods to expand the universe.

Writing Algorithms

LEAN Python algorithms inherit from QCAlgorithm:

from AlgorithmImports import *

class MyAlgo(QCAlgorithm):
    def Initialize(self):
        self.SetStartDate(2024, 1, 1)
        self.SetEndDate(2025, 1, 1)
        self.SetCash(100_000)
        self.AddEquity("SPY", Resolution.Daily)
        self.SetBenchmark("SPY")
        self.SetBrokerageModel(BrokerageName.InteractiveBrokersBrokerage,
                               AccountType.Margin)

    def OnData(self, data):
        if not self.Portfolio.Invested:
            self.SetHoldings("SPY", 1.0)

Key API patterns:

• self.History(symbol, periods, resolution) — get historical bars
• self.SetHoldings(symbol, weight) — target portfolio weight
• self.Liquidate(symbol) — close position
• self.AddUniverse(coarse_fn, fine_fn) — dynamic universe selection
• self.Schedule.On(date_rule, time_rule, action) — scheduled events
• self.Debug(msg) — log output

Analyzing Results

After a backtest run, check:

ls "$LEAN_ROOT/Results/"
# Key files: *-log.txt, *-order-log.txt, *.json (statistics)

Rebuild LEAN (if source changes)

export PATH="$PATH:$DOTNET_ROOT"
cd "$LEAN_ROOT"
dotnet build QuantConnect.Lean.sln -c Debug

Security Notes

Config.json Safety

The run_backtest.sh script does NOT modify your original config.json. Instead, it:

1. Reads the original config as a template (read-only)
2. Creates a separate config.backtest.json with only algorithm fields changed (class name, file path, language, environment=backtesting)
3. Temporarily swaps it in for the LEAN run, then restores the original via a trap cleanup handler

The configure_algo.py helper performs the field substitution in an isolated output file. Your original config — including any Interactive Brokers credentials for live trading — is never modified.

Modified fields (in the temp copy only):

• algorithm-type-name — set to the requested class name
• algorithm-language — set to Python
• algorithm-location — set to the requested .py file path
• environment — set to backtesting

Network Access

The setup instructions involve network downloads:

• git clone from GitHub (QuantConnect/Lean repository)
• dotnet build may restore NuGet packages
• pip install yfinance pandas installs Python packages from PyPI
• download_us_universe.py fetches market data from Yahoo Finance

All downloads are from well-known public sources. For maximum isolation, run setup in a container or VM.

Environment Variables

This skill requires the following environment variables at runtime:

• LEAN_ROOT — path to your cloned LEAN repository
• DOTNET_ROOT — path to your .NET SDK installation
• PYTHONNET_PYDLL — path to Python shared library (auto-detected from $LEAN_ROOT/.libs/ if not set)

These are declared in the skill metadata and must be set before use.

Troubleshooting

• "No data files found" → Check data-folder in config.json points to correct path
• Python import errors → LEAN bundles its own Python; check python-venv config if using custom packages
• Slow backtest → Reduce universe size or date range; check Resolution (Minute >> Daily)
• IB connection issues → Verify TWS/Gateway is running, port matches config (default 4002 for Gateway)
• LEAN_ROOT not set → Add export LEAN_ROOT="$HOME/lean" to your shell profile
• dotnet not found → Add export PATH="$PATH:$DOTNET_ROOT" to your shell profile
• Runtime.PythonDLL was not set → Set PYTHONNET_PYDLL to the Python shared library path (see env var table above)