a simple backtesting library
from functools import partial
import pandas as pd
from bt import Strategy, indicators
class SMA(Strategy):
"""
A Simple moving average strategy example
"""
def __init__(self, data: pd.DataFrame, start_index=None) -> None:
super().__init__(data, start_index)
sma_short = partial(sma, window=10)
self.add_indicator(sma_short, "sma_short")
sma_long = partial(sma, window=30)
self.add_indicator(sma_long, "sma_long")
def is_buy(self) -> bool:
return self.data.sma_short.iloc[-1] > self.data.sma_long.iloc[-1]
def is_sell(self) -> bool:
return self.data.sma_short.iloc[-1] < self.data.sma_long.iloc[-1]
if __name__ == "__main__":
df = pd.read_csv(
"../data/aapl.csv", index_col="date", parse_dates=True, infer_datetime_format=True
).drop(columns="Unnamed: 0")
# Initializing the Strategy object with the data passed to it.
strat = SMA(df)
# Backtest using the previously passed data
strat.backtest()
strat.plot_results(
"sma",
show=False,
plot_indicators=[["sma_short", "sma_long"]],
plot_table=True,
auto_open=True,
)
The strategy is an Abstract Base Class (ABC) that requires two methods to be defined within it's subclasses is_buy() and is_sell().
These two functions determine if, at any given point in time during the backtest, the strategy should buy or sell.
For example if one wanted to define a buy function to buy anytime the current price is higher than the previous tick's price you could do it like so
def is_buy(self):
# self.data refers to the currenly available data when backtesting
# or everything from the data passed in up to the current index
# `data_passed_in[: current_index]`
# Get the most recent price and the previous price
current_price = self.data['close'].iloc[-1]
prev_price = self.data['close'].iloc[-2]
if current_price > prev_price:
return True # Return True to indicate a buy signal
return False # Return False to not indicate a buy signal
The is_sell() function works in a very similar fashion. Return True if you wish to sell, and False otherwise.
def is_sell(self):
current_price = self.data['close'].iloc[-1]
prev_price = self.data['close'].iloc[-2]
if current_price > prev_price:
return True # Return True to sell
return False
