open-backtest 5.0.1

Open Backtest is a beginner friendly & powerful backtesting engine for crypto trading

Open Backtest

Open source & beginner friendly crypto trading backtest library

• French page

pip install open-backtest

Wanna contact me ? 👋

https://discord.gg/wfpGXvjj9t

Wanna support my work ? 💰

• paypal: [e-mail]
• usdt (ERC20): 0x17B516E9cA55C330B6b2bd2830042cAf5C7ecD7a
• btc: 34vo6zxSFYS5QJM6dpr4JLHVEo5vZ5owZH
• eth: 0xF7f87bc828707354AAfae235dE584F27bDCc9569

thanks if you do it 💖

What is it ? 📈

Passionate about the world of crypto and about development I decided to create a python library because I found very annoying for beginners to just run a simple backtest. Open Backtest got created to give apprentice but also confirmed programmers a powerful and easy to use backtesting tool

How does it work ? 🔧

Open Backtest is currently made with a core engine that use different classes, it can run a backtest with binance data and it can handle different timeframes. The library can also download and save data as a csv file to be able to load it to save a considerable amount of time. The wallet class will handle orders and the data handler will summarize and calculate all required data to analyze the backtest but also to plot graphs.

Requirements :

• Pandas
• Numpy
• Matplotlib
• Python-binance

All requirements will be downloaded and installed with Open Backtest installation

Doc 📝

The documentation will be divided in two parts, at the moment two engines are made but I want to add a lot of engines. The first part of the doc will show how to run a backtest. The second part of the doc will describe more technically the classes and functions that can be used

How to run a backtest ?

We will see here an example that use the first engine

# ------------------------------------------------------------------
# First of all, let's see all our imports
# ------------------------------------------------------------------

# The library Technical Analyse is already included when installing open backtest
# it allow to add a lot of indicators very useful for trading strategy
from ta import trend, momentum

# Let's import here 3 classes of Open Backtest we will later see how to use it
from OpenBacktest.ObtEngine import Engine, Container, Pair

# Python Binance is also included with Open Backtest it allow us to get the market data
from binance.client import Client

# ------------------------------------------------------------------
# The let's initialise our classes
# ------------------------------------------------------------------

# First of all we are here creating our container, it will contain all of our market pairs
container = Container()

# This is the tricky part ! We will here add to the container pairs we want to use by 2 functions,
# container.add_main_pair() and container.add_pair(). These 2 functions will have the same parameter ! The main pair
# will be the pair with the timeframe used to run your backtest /!\ it's required. There is also an other function
# named container.add_pair(), it will be used to add some other pairs that are not required. Note that it's useless
# for this engine to add pairs of others symbols, I mean if your main pair is for example Ethereum Usdt it's useless
# to add Bitcoin Usdt but it's technically possible. The real interest is to add the same pair but with a different
# timeframe ! Follow the structure below to add your pairs. The parameter of both functions is a Pair class with 5
# parameters, the parameters are quite self-explanatory but just to clarify, name is just a recognizable name for you
# that will be used later to get the data of a specific pair, the path is the location of files that already exist or
# the location of futures files that will be saved, this parameter is optional.

# We register here our main pair ! The data will be get for the pair Ethereum - Usdt from the 01 january 2021 to now
# with candles of 1 hour
container.add_main_pair(
    Pair(market_pair="ETHUSDT", start="01 january 2021", timeframe=Client.KLINE_INTERVAL_1HOUR, name="ETHUSDT",
         path=""))
# We register here a second pair with a largest timeframe ! The data will be get for the pair Ethereum - Usdt from
# the 01 january 2021 to now with candles of 1 day, note that the name is not the same than our first pair !
container.add_pair(
    Pair(market_pair="ETHUSDT", start="01 january 2021", timeframe=Client.KLINE_INTERVAL_1DAY, name="ETHUSDT1d",
         path=""))

# Let's now initialise our engine with our container
engine = Engine(container)

# This line is not required ! it's used to save our data as csv files to be able to just load it for the next backtest
container.save_all(default_path="")

# We are here enriching our dataframes with technical indicators using TA lib more information here
# https://technical-analysis-library-in-python.readthedocs.io/en/latest/

# Let's add to our main dataframe 2 EMA
engine.main_dataframe()["EMA3"] = trend.ema_indicator(engine.main_dataframe()['close'], 3)
engine.main_dataframe()["EMA100"] = trend.ema_indicator(engine.main_dataframe()['close'], 100)

# Let's add to our second dataframe 2 EMA, to get our second dataframe we will now use our names that we
# configured before !
engine.get_sub_dataframe("ETHUSDT1d")["EMA3"] = trend.ema_indicator(engine.get_sub_dataframe("ETHUSDT1d")["close"], 3)
engine.get_sub_dataframe("ETHUSDT1d")["EMA100"] = trend.ema_indicator(engine.get_sub_dataframe("ETHUSDT1d")["close"],
                                                                      100)


# We will now set a buy condition that will return True when it want to buy and same for a sell condition. The engine
# will call this function with the main dataframe and the current index
def buy_condition(dataframe, index):
    if dataframe["EMA3"][index] >= dataframe["EMA100"][index]:
        return True

    # For these simple buy & sell conditions we will don't use our second dataframe but if you want to get data of it
    # it's simple !
    # The Pair class
    second_pair = engine.container.get_pair("ETHUSDT1d")
    # The dataframe
    second_dataframe = second_pair.dataframe
    # And the index you have to use to get your data corresponding to the current index of the main dataframe
    second_index = second_pair.get_index(dataframe["timestamp"][index])
    # So we can now access the ema for your timeframe of 1d like this :
    current_value_of_ema = second_dataframe["EMA3"][second_index]

    # You can also place take profit and stop loss !
    engine.set_take_profit(index, percent_target=50)
    engine.set_stop_loss(index, percent_target=-50)
    # Our bot will now sell all of it tokens when the price will increase or decrease by 50%


# Same here as sell condition
def sell_condition(dataframe, index):
    if dataframe["EMA3"][index] <= dataframe["EMA100"][index]:
        return True


# ------------------------------------------------------------------
# The let's run our backtest !
# ------------------------------------------------------------------

# This function is used to register our conditions
engine.register_sell_and_buy_condition(buy_condition, sell_condition)

# This function is used to run the backtest, first parameter is the coin name, second is the token name, third
# is your initial coin balance 4th is your initial token balance 5th is your taker fees and 6th is your maker fees
engine.run_sell_and_buy_condition("USDT", "Ethereum", 20, 0, 0.065, 0.019)

# We use this function to summarize and display the result of our backtest
engine.wallet.data_handler.display_wallet()

# And we finally use it to plot graphs of price and balance evolution, the parameter is not required and is the size
# of the sell and buy point on price graph
engine.wallet.data_handler.plot_wallet(25)

# -----------------------------------------------------------------------------------------------------------------
# And that's finish ! Hope you like and that it wasn't hard ! If you have any question dm me on discord: Shaft#3796
# -----------------------------------------------------------------------------------------------------------------

We will now see an example that use the asymmetric engine

# -----------------------------------------------------------------------------------------------------------------
# First of all, let's see all our imports but before, important message !!
#
# Read and understand first how does the Symmetric engine work before to try with this engine,
# it's important because I will not show here some part of the library already explained with the symmetric engine
#
# -----------------------------------------------------------------------------------------------------------------

# The library Technical Analyse is already included when installing open backtest
# it allow to add a lot of indicators very useful for trading strategy
from ta import trend, momentum

# Let's import here 4 classes of Open Backtest we will later see how to use it
from OpenBacktest.ObtEngine import AsymmetricEngine, Container, Pair, Report

# Python Binance is also included with Open Backtest it allow us to get the market data
from binance.client import Client

# ------------------------------------------------------------------
# The let's initialise our classes
# ------------------------------------------------------------------

# First of all we are here creating our container, it will contain all of our market pairs
container = Container()

# Like for a symmetric engine let's register our main pair but this time I will not show you how to register
# others pairs with others timeframe but it's possible as explained for the symmetric engine
container.add_main_pair(
    Pair(market_pair="ETHUSDT", start="01 january 2021", timeframe=Client.KLINE_INTERVAL_1HOUR, name="ETHUSDT",
         path=""))

# Let's now initialise our engine with our container but this time, an assymetric engine
engine = AsymmetricEngine(container)

# We are here enriching our dataframes with technical indicators using TA lib more information here
# https://technical-analysis-library-in-python.readthedocs.io/en/latest/

# Let's add to our main dataframe 2 EMA
engine.main_dataframe()["EMA3"] = trend.ema_indicator(engine.main_dataframe()['close'], 3)
engine.main_dataframe()["EMA100"] = trend.ema_indicator(engine.main_dataframe()['close'], 100)


# This time our Asymmetric engine will just work with a strategy function that will return a report !
def strategy(dataframe, index):
    # we will with this function return a report, return None to do nothing or return a Report class to pass an order,
    # the first parameter of our report is required and will be the order type, "sell" or "buy". The second parameter
    # is not required and is the amount of token or coin you want to spend. The third parameter is also not required
    # and is the amount in percent of your wallet you want to spend
    if dataframe["EMA3"][index] >= dataframe["EMA100"][index]:
        return Report("buy", percent_amount=50)
    if dataframe["EMA3"][index] <= dataframe["EMA100"][index]:
        return Report("sell", percent_amount=50)

    # you can also here use take profit and stop loss as showed with the symmetric engine !
    # note that with this simple strategy I'm not fully using the asymmetric engine. This engine can be used for
    # more advanced strategies for example grid trading


# ------------------------------------------------------------------
# The let's run our backtest !
# ------------------------------------------------------------------

# This function is used to register our strategy
engine.register_strategy(strategy)

# This function is used to run the backtest, first parameter is the coin name, second is the token name, third
# is your initial coin balance 4th is your initial token balance 5th is your taker fees and 6th is your maker fees
engine.run_strategy("USDT", "Ethereum", 20, 0, 0.065, 0.019)

# We use this function to summarize and display the result of our backtest
engine.wallet.data_handler.display_wallet()

# And we finally use it to plot graphs of price and balance evolution, the parameter is not required and is the size
# of the sell and buy point on price graph
engine.wallet.data_handler.plot_wallet(25)

# -----------------------------------------------------------------------------------------------------------------
# And that's finish ! Hope you like and that it wasn't hard ! If you have any question dm me on discord: Shaft#3796
# -----------------------------------------------------------------------------------------------------------------

Next part is coming soon