market-break

a set of utilities for backtesting vectorized trading algorithms, high-frequency trading strategies and investing strategies.

Installation

pip install python-market-break

Connect to a live database for real-time data.

from sqlalchemy import create_engine

engine = create_engine('sqlite:///database/2024-06-14.sqlite')

Extract The data from the database into a DataFrame.

from cryptocore.market.database import extract_dataframe, table_name

EXCHANGE = "binance"
SYMBOL = "ETH/USDT"

df = extract_dataframe(engine, table_name(exchange=EXCHANGE, symbol=SYMBOL))
df = df.iloc[3500:7000].reset_index()

df

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

Import column names to handle the data.

from market_break.labels import BID, ASK, DATETIME, ENTRY, EXIT, TYPE, LONG

Import a class of backtesting functions.

from market_break.backtest import Trades

Generate trades record, both short and long as a DataFrame, from definition of up-trends and down-trends.

trades = Trades.generate(
    up=df[ASK] > df[ASK].shift(1),
    down=df[BID] < df[BID].shift(1),
    adjust=True
)

Process the results of the generated trades.

FEE = 0.001

returns = Trades.returns(trades, bid=df[BID], ask=df[ASK], fee=FEE)

Imports a class of Plotting trading results.

from market_break.backtest import Plot

Plot a histogram of the trades returns.

Plot.returns_histogram(returns, bins=35)

Plot a graph of the trades returns.

Plot.returns_signals(returns, index=df[DATETIME].iloc)

Plot pie graphs of the winning and losing trades and their profits and losses.

Plot.returns_pie(returns)

Plot the signals for long and short entries and exits, with the bid-ask spread. Also plotting The balance, profits and losses.

long_trades = trades[trades[TYPE] == LONG]

Plot.price_signals(
    bid=df[BID], ask=df[ASK], index=df[DATETIME].iloc,
    long=long_trades[ENTRY], short=long_trades[EXIT]
)
Plot.returns_balance(returns, index=df[DATETIME].iloc)

Import a class for generating a report from the results.

from market_break.backtest import Report

Generating and displaying the results with the report.

report = Report.generate(
    index=df[DATETIME], returns=returns, 
    long=trades[EXIT].values, short=trades[EXIT].values
)

print(Report.repr(report))

[Index]
start                   2024-06-14 11:19:38.20
end                     2024-06-14 12:18:00.02
total duration              0 days 00:58:21.82
min. tick duration                  0:00:00.06
max. tick duration                  0:00:01.90
avg. tick duration                  0:00:01.00
ticks                                     3500

[Trades]
long trades                                125
short trades                               125
min. TUW                                   0:0
max. TUW                                   0:0
avg. TUW                                   0:0

[Gains]
[%] min. gain                           0.0003
[%] max. gain                           0.1482
[%] avg. gain                           0.0291
[%] total gains                         2.4411
winning trades                              84

[Losses]
[%] min. loss                          -0.0003
[%] max. loss                           -0.017
[%] avg. loss                          -0.0063
[%] total losses                       -0.2383
losing trades                               38

[Performance]
PnL factor                              10.245
avg. profit factor                      1.0002
[%] win rate                              67.2
[%] total profit                        2.2029