Zipline is a financial backtester for trading algorithms written in Python. The system is fundamentally event-driven and a close approximation of how live-trading systems operate.
Zipline is currently used in production as the backtesting engine powering Quantopian (https://www.quantopian.com) -- a free, community-centered platform that allows development and real-time backtesting of trading algorithms in the web browser.
Want to contribute? See our open requests and our general guidelines below.
Discussion of the project is held at the Google Group, [email protected], https://groups.google.com/forum/#!forum/zipline.
-
Ease of use: Zipline tries to get out of your way so that you can focus on algorithm development. See below for a code example.
-
Zipline comes "batteries included" as many common statistics like moving average and linear regression can be readily accessed from within a user-written algorithm.
-
Input of historical data and output of performance statistics is based on Pandas DataFrames to integrate nicely into the existing Python eco-system.
-
Statistic and machine learning libraries like matplotlib, scipy, statsmodels, and sklearn support development, analysis and visualization of state-of-the-art trading systems.
Since zipline is pure-python code it should be very easy to install and set up with pip:
pip install zipline
If there are problems installing the dependencies or zipline we recommend installing these packages via some other means. For Windows, the Enthought Python Distribution includes most of the necessary dependencies. On OSX, the Scipy Superpack works very well.
- Python (>= 2.7.2)
- numpy (>= 1.6.0)
- pandas (>= 0.9.0)
- pytz
- msgpack-python
- Logbook
- blist
The following code implements a simple dual moving average algorithm and tests it on data extracted from yahoo finance.
from zipline.algorithm import TradingAlgorithm
from zipline.transforms import MovingAverage
from zipline.utils.factory import load_from_yahoo
class DualMovingAverage(TradingAlgorithm):
"""Dual Moving Average algorithm.
"""
def initialize(self, short_window=200, long_window=400):
# Add 2 mavg transforms, one with a long window, one
# with a short window.
self.add_transform(MovingAverage, 'short_mavg', ['price'],
market_aware=True,
window_length=short_window)
self.add_transform(MovingAverage, 'long_mavg', ['price'],
market_aware=True,
window_length=long_window)
# To keep track of whether we invested in the stock or not
self.invested = False
self.short_mavg = []
self.long_mavg = []
def handle_data(self, data):
if (data['AAPL'].short_mavg['price'] > data['AAPL'].long_mavg['price']) and not self.invested:
self.order('AAPL', 100)
self.invested = True
elif (data['AAPL'].short_mavg['price'] < data['AAPL'].long_mavg['price']) and self.invested:
self.order('AAPL', -100)
self.invested = False
# Save mavgs for later analysis.
self.short_mavg.append(data['AAPL'].short_mavg['price'])
self.long_mavg.append(data['AAPL'].long_mavg['price'])
data = load_from_yahoo()
dma = DualMovingAverage()
results = dma.run(data)
You can find other examples in the zipline/examples directory.
If you would like to contribute, please see our Contribution Requests: https://github.com/quantopian/zipline/wiki/Contribution-Requests
Thank you for all the help so far!
- @snth
- @yinhm for integrating zipline with @yinhm/datafeed
- Jeremiah Lowin for teaching us the nuances of Sharpe and Sortino Ratios
- Quantopian Team
(alert us if we've inadvertantly missed listing you here!)
To ensure that changes and patches are focused on behavior changes, the zipline codebase adheres to both PEP-8, http://www.python.org/dev/peps/pep-0008/, and pyflakes, https://launchpad.net/pyflakes/.
The maintainers check the code using the flake8 script, https://github.com/bmcustodio/flake8, which is included in the requirements_dev.txt.
Before submitting patches or pull requests, please ensure that your
changes pass flake8 zipline tests
The source for Zipline is hosted at https://github.com/quantopian/zipline.
For other questions, please contact [email protected].