Quick Start Guide

Get up and running with RLX Backtester in minutes

1 Installation

Install RLX Backtester using pip. The package includes both the Python API and the pre-compiled Rust core.

Terminal

pip install rlxbt

Python Version

RLX requires Python 3.10 or higher. We recommend using a virtual environment.

Verify the installation:

Python

import rlxbt
print(rlxbt.__version__)

2 Your First Backtest

Let's run a simple backtest using pre-calculated signals. This is the fastest way to test RLX.

Python

import pandas as pd
from rlxbt import TradingEngine

# Load your OHLCV data with signals
df = pd.read_csv("data.csv")

# Ensure timestamp is in Unix seconds
df['timestamp'] = pd.to_datetime(df['time']).astype('int64') // 10**9

# Create the trading engine
engine = TradingEngine(
    initial_capital=100000.0,
    commission=0.001,  # 0.1% commission
    slippage=0.0
)

# Run backtest (signal column: 1=Long, -1=Short, 0=Flat)
result = engine.run_with_signals(df, signal_column="signal")

# View results
print(f"Total Return: {result.total_return:.2%}")
print(f"Sharpe Ratio: {result.metrics['sharpe_ratio']:.2f}")
print(f"Max Drawdown: {result.metrics['max_drawdown']:.2%}")
print(f"Total Trades: {result.total_trades}")

⚡

Performance

RLX processes ~4.5 million bars per second with full event-driven simulation.

3 Data Format

RLX expects OHLCV data with specific columns. Here's the required format:

Column	Type	Description
timestamp	int	Unix timestamp in seconds
open	float	Opening price
high	float	Highest price
low	float	Lowest price
close	float	Closing price
volume	float	Trading volume

Example CSV:

data.csv

timestamp,open,high,low,close,volume,signal
1609459200,29000.0,29500.0,28800.0,29200.0,1000.0,1
1609462800,29200.0,29300.0,29000.0,29100.0,800.0,1
1609466400,29100.0,29400.0,29050.0,29350.0,1200.0,0
...

4 License Activation

For production use, you need a valid license key. Get yours at /pricing.

Python

from rlxbt import TradingEngine

# Pass license key directly
engine = TradingEngine(
    initial_capital=100000.0,
    license_key="rlx_pro_xxxxxxxxxxxx_xxxx"
)

# Or set via environment variable
# export RLX_LICENSE_KEY="rlx_pro_xxxxxxxxxxxx_xxxx"

Development Mode

Builds compiled with --features offline_license don't require a license key.

5 Creating Strategies

Create a custom strategy by subclassing Strategy:

Python

import pandas as pd
from rlxbt import Strategy, Backtester

class SMAStrategy(Strategy):
    def __init__(self, short_period=10, long_period=50):
        super().__init__()
        self.short = short_period
        self.long = long_period

    def generate_signals(self, data: pd.DataFrame) -> pd.DataFrame:
        # Calculate moving averages
        short_ma = data['close'].rolling(self.short).mean()
        long_ma = data['close'].rolling(self.long).mean()
        
        # Generate signals
        signals = pd.DataFrame(index=data.index)
        signals['signal'] = 0
        signals.loc[short_ma > long_ma, 'signal'] = 1   # Long
        signals.loc[short_ma < long_ma, 'signal'] = -1  # Short
        
        return signals

# Run backtest
strategy = SMAStrategy(short_period=20, long_period=50)
backtester = Backtester(initial_capital=100000.0)
result = backtester.run(strategy, df)

6 Understanding Results

The BacktestResult object contains comprehensive metrics:

Basic Metrics

total_return — Total percentage return
total_trades — Number of completed trades
winning_trades — Profitable trades count
losing_trades — Losing trades count

Risk Metrics

sharpe_ratio — Risk-adjusted return
sortino_ratio — Downside risk-adjusted
max_drawdown — Peak-to-trough decline
calmar_ratio — Return / Max drawdown

Python

# Access all metrics
for key, value in result.metrics.items():
    print(f"{key}: {value}")

# Access trade history
for trade in result.trades[:5]:
    print(f"Entry: {trade.entry_time}, PnL: ${trade.pnl:.2f}")

# Equity curve for visualization
equity = result.equity_curve  # List[float]
timestamps = result.equity_curve_timestamps  # List[int]