backintime

Tool for testing trading strategies on historical data


Keywords
backtesting, backtesting-engine, backtesting-trading-strategies, crypto, trading
License
MIT
Install
pip install backintime==1.6.2

Documentation

backintime 1.6.1 (β)

A framework for trading strategies backtesting with Python

Trailing stop, stop limit and OCO orders are not supported as of the current version.
Expected in 1.x.x releases.

Features

  • Market/Limit orders management
  • Use CSV or Binance API as a data source
  • The same data can be represented in up to 16 timeframes
    (few short candles is compressed to longer one)
  • Brief trading history statistics (win rate, avg. profit, etc.)
  • Export trades to csv

This is how it looks like - MACD strategy

see macd strategy explained

from backintime import TradingStrategy, Timeframes
from backintime.oscillators.macd import macd
'''
Extend TradingStrategy class and implement __call__ method
to have your own strategy
'''
class MacdStrategy(TradingStrategy):
    # declare required oscillators here for later use
    using_oscillators = ( macd(Timeframes.H4), )

    def __call__(self):
        # runs each time a new candle closes
        macd = self.oscillators.get('MACD_H4')

        if not self.position and macd.crossover_up():
            self._buy()     # buy at market

        elif self.position and macd.crossover_down():
            self._sell()    # sell at market

backtesting is done as follows (with binance API data):

# add the following import to the ones above
from backintime import BinanceApiCandles

feed = BinanceApiCandles('BTCUSDT', Timeframes.H4)
backtester = Backtester(MacdStrategy, feed)

backtester.run_test(since='2020-01-01', start_money=10000)
# the result is available as a printable instance
res = backtester.results()
print(res)
# and also can be saved to a csv file
res.to_csv('filename.csv', sep=';', summary=True)

Alternatively, you can use a csv file on your local machine as source

from backintime import TimeframeDump, TimeframeDumpScheme
# specify column indexes in input csv
columns = TimeframeDumpScheme(
    open_time=0, close_time=6,
    open=1, high=3, low=4,
    close=2, volume=5)

feed = TimeframeDump('h4.csv', Timeframes.H4, columns)
backtester = Backtester(MacdStrategy, feed)
backtester.run_test('2020-01-01', 10000)
print(backtester.results())

Install

pip install backintime

License

MIT

Author

Akim Mukhtarov @akim_int80h