aljaz/hummingbot-dashboard
1
0
Fork 0
mirror of https://github.com/aljazceru/hummingbot-dashboard.git synced 2026-01-29 10:04:20 +01:00
Code Issues Packages Projects Releases Wiki Activity

(feat) remove strategy optimizer

Browse Source
This commit is contained in:
cardosofede
2023-07-20 15:47:18 +02:00
parent 437613ab55
commit f81579fdca
1 changed files with 0 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files

77
quants_lab/strategy/strategy_optimizer.py
View File

@@ -1,77 +0,0 @@
import optuna
from quants_lab.backtesting.backtesting import Backtesting
from quants_lab.backtesting.backtesting_analysis import BacktestingAnalysis
from quants_lab.strategy.mean_reversion.bollinger import Bollinger
from quants_lab.strategy.mean_reversion.macd_bb import MACDBB
from optuna.exceptions import TrialPruned
from quants_lab.strategy.mean_reversion.stat_arb import StatArb
STUDY_NAME = "stat_arb"
def objective(trial):
# strategy = Bollinger(
# exchange="binance_perpetual",
# trading_pair="ETH-USDT",
# interval="3m",
# bb_length=trial.suggest_int("bb_length", 20, 300),
# bb_std=trial.suggest_float("bb_std", 1.0, 3.0),
# bb_long_threshold=trial.suggest_float("bb_long_threshold", -0.5, 0.3),
# bb_short_threshold=trial.suggest_float("bb_short_threshold", 0.7, 1.5),
# )
# fast_macd = trial.suggest_int("fast_macd", 10, 50)
# strategy = MACDBB(
# exchange="binance_perpetual",
# trading_pair="ETH-USDT",
# interval="3m",
# bb_length=trial.suggest_int("bb_length", 20, 300),
# bb_std=trial.suggest_float("bb_std", 1.0, 3.0),
# bb_long_threshold=trial.suggest_float("bb_long_threshold", -0.5, 0.3),
# bb_short_threshold=trial.suggest_float("bb_short_threshold", 0.7, 1.5),
# fast_macd=fast_macd,
# slow_macd=trial.suggest_int("slow_macd", fast_macd + 1, 100),
# signal_macd=trial.suggest_int("signal_macd", 8, 54)
# )
strategy = StatArb(trading_pair="ETH-USDT",
periods=trial.suggest_int("periods", 10, 150),
deviation_threshold=trial.suggest_float("deviation_threshold", 0.9, 2.0))
try:
backtesting = Backtesting(strategy=strategy)
backtesting_result = backtesting.run_backtesting(
start='2021-04-01',
order_amount=50,
leverage=20,
initial_portfolio=100,
take_profit_multiplier=trial.suggest_float("take_profit_multiplier", 1.0, 3.0),
stop_loss_multiplier=trial.suggest_float("stop_loss_multiplier", 1.0, 3.0),
time_limit=60 * 60 * trial.suggest_int("time_limit", 1, 24),
std_span=None,
)
backtesting_analysis = BacktestingAnalysis(
positions=backtesting_result,
)
trial.set_user_attr("net_profit_usd", backtesting_analysis.net_profit_usd())
trial.set_user_attr("net_profit_pct", backtesting_analysis.net_profit_pct())
trial.set_user_attr("max_drawdown_usd", backtesting_analysis.max_drawdown_usd())
trial.set_user_attr("max_drawdown_pct", backtesting_analysis.max_drawdown_pct())
trial.set_user_attr("sharpe_ratio", backtesting_analysis.sharpe_ratio())
trial.set_user_attr("accuracy", backtesting_analysis.accuracy())
trial.set_user_attr("total_positions", backtesting_analysis.total_positions())
trial.set_user_attr("win_signals", backtesting_analysis.win_signals().shape[0])
trial.set_user_attr("loss_signals", backtesting_analysis.loss_signals().shape[0])
trial.set_user_attr("profit_factor", backtesting_analysis.profit_factor())
trial.set_user_attr("duration_in_hours", backtesting_analysis.duration_in_minutes() / 60)
trial.set_user_attr("avg_trading_time_in_hours", backtesting_analysis.avg_trading_time_in_minutes() / 60)
return backtesting_analysis.net_profit_pct()
except Exception as e:
print(e)
raise TrialPruned()
study = optuna.create_study(direction="maximize", study_name=STUDY_NAME, storage="sqlite:///backtesting_report.db",
load_if_exists=True)
study.optimize(objective, n_trials=2000)