Skip to main content

​
Introduction

GlowBack provides high-performance Python bindings built with PyO3, enabling you to leverage Rust’s speed and safety while writing backtesting strategies in Python. The API is designed to be intuitive for Python developers while maintaining the performance characteristics of the underlying Rust engine.

​
Installation

Install GlowBack using pip: 
pip install glowback
For development installations: 
git clone https://github.com/yourorg/glowback.git
cd glowback
pip install -e .

​
Quick start

Here’s a simple example to get started with GlowBack: 
from glowback import BacktestEngine, run_buy_and_hold

# Run a simple buy-and-hold strategy
result = run_buy_and_hold(
    symbols=["AAPL"],
    start_date="2020-01-01T00:00:00Z",
    end_date="2023-12-31T23:59:59Z",
    resolution="day",
    initial_capital=100000.0,
    name="My First Backtest"
)

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

# Plot equity curve
result.plot_equity(show=True)

​
Core classes

The GlowBack Python API consists of the following main components:

DataManager

Load and manage market data from multiple sources

BacktestEngine

Execute backtests with custom strategies

Strategy Classes

Define and implement trading strategies

Data Types

Symbols, bars, and portfolio types

​
API conventions

​
Date formats

All dates must be in RFC3339/ISO8601 format with timezone information: 
# Correct formats
start_date = "2020-01-01T00:00:00Z"           # UTC
start_date = "2020-01-01T00:00:00-05:00"      # EST
start_date = "2020-01-01T00:00:00+00:00"      # UTC with offset

# Incorrect formats (will raise ValueError)
start_date = "2020-01-01"                      # Missing time and timezone
start_date = "01/01/2020"                      # Wrong format

​
Resolutions

GlowBack supports the following time resolutions:
  • "minute" or "1m" - 1-minute bars
  • "hour" or "1h" - 1-hour bars
  • "day" or "1d" - Daily bars (default)
engine = BacktestEngine(
    symbols=["AAPL"],
    start_date="2020-01-01T00:00:00Z",
    end_date="2023-12-31T23:59:59Z",
    resolution="day"  # or "1d", "hour", "1h", "minute", "1m"
)

​
Asset classes

When creating symbols, you can specify the asset class:
  • "equity" - Stocks (default)
  • "crypto" - Cryptocurrencies
  • "forex" - Foreign exchange
  • "commodity" - Commodities
  • "bond" - Bonds
from glowback import Symbol

# Create symbols with different asset classes
aapl = Symbol("AAPL", "NASDAQ", "equity")
btc = Symbol("BTC", "BINANCE", "crypto")
eurusd = Symbol("EURUSD", "FOREX", "forex")

​
Error handling

GlowBack raises standard Python exceptions: 
from glowback import BacktestEngine

try:
    engine = BacktestEngine(
        symbols=[],  # Empty list
        start_date="2020-01-01T00:00:00Z",
        end_date="2023-12-31T23:59:59Z"
    )
except ValueError as e:
    print(f"Configuration error: {e}")
    # Output: Configuration error: symbols list cannot be empty

try:
    engine = BacktestEngine(
        symbols=["AAPL"],
        start_date="invalid-date",
        end_date="2023-12-31T23:59:59Z"
    )
except ValueError as e:
    print(f"Date parsing error: {e}")
    # Output: Date parsing error: Invalid start_date format: ...

​
Working with results

Backtest results provide multiple ways to access and visualize data: 
result = run_buy_and_hold(
    symbols=["AAPL"],
    start_date="2020-01-01T00:00:00Z",
    end_date="2023-12-31T23:59:59Z"
)

# Access metrics as dictionary
metrics = result.metrics_summary
print(f"Total Return: {metrics['total_return']:.2f}%")

# Get equity curve as list of dicts
equity_data = result.equity_curve
for point in equity_data[:5]:  # First 5 points
    print(f"{point['timestamp']}: ${point['value']:.2f}")

# Convert to pandas DataFrame (requires pandas)
df = result.to_dataframe(index='timestamp')
print(df.head())

# Get metrics as DataFrame
metrics_df = result.metrics_dataframe()
print(metrics_df)

​
Jupyter notebook integration

GlowBack is designed to work seamlessly in Jupyter notebooks: 
import glowback

# Results automatically display in notebooks
result = glowback.run_buy_and_hold(
    symbols=["AAPL", "MSFT"],
    start_date="2020-01-01T00:00:00Z",
    end_date="2023-12-31T23:59:59Z",
    initial_capital=100000.0
)

# Show summary with plot
result.summary(plot=True)

# Or just plot equity curve
ax = result.plot_equity(show=False)
ax.set_title("My Custom Title")

​
Next steps

Load market data

Learn how to load data from various sources

Run backtests

Execute backtests with the BacktestEngine

Custom strategies

Create your own trading strategies

API examples

View complete working examples

Build docs developers (and LLMs) love

Get started for freeTalk to us