adelorenzo/stable
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

integrate hyperopt and implement subcommand

Browse Source
This commit is contained in:
gcarq
2017-11-25 01:04:11 +01:00
parent 7fa5846c6b
commit b9c4eafd96
9 changed files with 191 additions and 167 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
freqtrade/optimize/__init__.py
View File

@@ -1 +1,41 @@
from . import backtesting
# pragma pylint: disable=missing-docstring
import json
import os
from typing import Optional, List, Dict
from pandas import DataFrame
from freqtrade.analyze import populate_indicators, parse_ticker_dataframe
def load_data(ticker_interval: int = 5, pairs: Optional[List[str]] = None) -> Dict[str, List]:
"""
Loads ticker history data for the given parameters
:param ticker_interval: ticker interval in minutes
:param pairs: list of pairs
:return: dict
"""
path = os.path.abspath(os.path.dirname(__file__))
result = {}
_pairs = pairs or [
'BTC_BCC', 'BTC_ETH', 'BTC_DASH', 'BTC_POWR', 'BTC_ETC',
'BTC_VTC', 'BTC_WAVES', 'BTC_LSK', 'BTC_XLM', 'BTC_OK',
]
for pair in _pairs:
with open('{abspath}/../tests/testdata/{pair}-{ticker_interval}.json'.format(
abspath=path,
pair=pair,
ticker_interval=ticker_interval,
)) as tickerdata:
result[pair] = json.load(tickerdata)
return result
def preprocess(tickerdata: Dict[str, List]) -> Dict[str, DataFrame]:
"""Creates a dataframe and populates indicators for given ticker data"""
processed = {}
for pair, pair_data in tickerdata.items():
processed[pair] = populate_indicators(parse_ticker_dataframe(pair_data))
return processed

25
freqtrade/optimize/backtesting.py
View File

@@ -9,34 +9,17 @@ from pandas import DataFrame
from tabulate import tabulate
from freqtrade import exchange
from freqtrade.analyze import parse_ticker_dataframe, populate_indicators, \
populate_buy_trend, populate_sell_trend
from freqtrade.analyze import populate_buy_trend, populate_sell_trend
from freqtrade.exchange import Bittrex
from freqtrade.main import min_roi_reached
from freqtrade.misc import load_config
from freqtrade.optimize import load_data, preprocess
from freqtrade.persistence import Trade
from freqtrade.tests import load_backtesting_data
logger = logging.getLogger(__name__)
def format_results(results: DataFrame):
return ('Made {:6d} buys. Average profit {: 5.2f}%. '
'Total profit was {: 7.3f}. Average duration {:5.1f} mins.').format(
len(results.index),
results.profit.mean() * 100.0,
results.profit.sum(),
results.duration.mean() * 5,
)
def preprocess(backdata) -> Dict[str, DataFrame]:
processed = {}
for pair, pair_data in backdata.items():
processed[pair] = populate_indicators(parse_ticker_dataframe(pair_data))
return processed
def get_timeframe(data: Dict[str, Dict]) -> Tuple[arrow.Arrow, arrow.Arrow]:
"""
Get the maximum timeframe for the given backtest data
@@ -151,7 +134,7 @@ def start(args):
data[pair] = exchange.get_ticker_history(pair, args.ticker_interval)
else:
print('Using local backtesting data (ignoring whitelist in given config)...')
data = load_backtesting_data(args.ticker_interval)
data = load_data(args.ticker_interval)
print('Using stake_currency: {} ...\nUsing stake_amount: {} ...'.format(
config['stake_currency'], config['stake_amount']

199
freqtrade/optimize/hyperopt.py
View File

@@ -1,29 +1,124 @@
# pragma pylint: disable=missing-docstring,W0212
import logging
import os
from functools import reduce
from math import exp
from operator import itemgetter
import pytest
from hyperopt import fmin, tpe, hp, Trials, STATUS_OK
from pandas import DataFrame
from freqtrade import exchange
from freqtrade import exchange, optimize
from freqtrade.exchange import Bittrex
from freqtrade.optimize.backtesting import backtest, format_results
from freqtrade.optimize.backtesting import preprocess
from freqtrade.tests import load_backtesting_data
from freqtrade.optimize.backtesting import backtest
from freqtrade.vendor.qtpylib.indicators import crossed_above
logging.disable(logging.DEBUG) # disable debug logs that slow backtesting a lot
# set TARGET_TRADES to suit your number concurrent trades so its realistic to 20days of data
TARGET_TRADES = 1100
TOTAL_TRIES = 4
# pylint: disable=C0103
current_tries = 0
# Configuration and data used by hyperopt
PROCESSED = optimize.preprocess(optimize.load_data())
OPTIMIZE_CONFIG = {
'max_open_trades': 3,
'stake_currency': 'BTC',
'stake_amount': 0.01,
'minimal_roi': {
'40': 0.0,
'30': 0.01,
'20': 0.02,
'0': 0.04,
},
'stoploss': -0.10,
}
SPACE = {
'mfi': hp.choice('mfi', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('mfi-value', 5, 25, 1)}
]),
'fastd': hp.choice('fastd', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('fastd-value', 10, 50, 1)}
]),
'adx': hp.choice('adx', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('adx-value', 15, 50, 1)}
]),
'rsi': hp.choice('rsi', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('rsi-value', 20, 40, 1)}
]),
'uptrend_long_ema': hp.choice('uptrend_long_ema', [
{'enabled': False},
{'enabled': True}
]),
'uptrend_short_ema': hp.choice('uptrend_short_ema', [
{'enabled': False},
{'enabled': True}
]),
'over_sar': hp.choice('over_sar', [
{'enabled': False},
{'enabled': True}
]),
'green_candle': hp.choice('green_candle', [
{'enabled': False},
{'enabled': True}
]),
'uptrend_sma': hp.choice('uptrend_sma', [
{'enabled': False},
{'enabled': True}
]),
'trigger': hp.choice('trigger', [
{'type': 'lower_bb'},
{'type': 'faststoch10'},
{'type': 'ao_cross_zero'},
{'type': 'ema5_cross_ema10'},
{'type': 'macd_cross_signal'},
{'type': 'sar_reversal'},
{'type': 'stochf_cross'},
{'type': 'ht_sine'},
]),
}
def optimizer(params):
from freqtrade.optimize import backtesting
backtesting.populate_buy_trend = buy_strategy_generator(params)
results = backtest(OPTIMIZE_CONFIG, PROCESSED)
result = format_results(results)
total_profit = results.profit.sum() * 1000
trade_count = len(results.index)
trade_loss = 1 - 0.35 * exp(-(trade_count - TARGET_TRADES) ** 2 / 10 ** 5.2)
profit_loss = max(0, 1 - total_profit / 10000) # max profit 10000
# pylint: disable=W0603
global current_tries
current_tries += 1
print('{:5d}/{}: {}'.format(current_tries, TOTAL_TRIES, result))
return {
'loss': trade_loss + profit_loss,
'status': STATUS_OK,
'result': result
}
def format_results(results: DataFrame):
return ('Made {:6d} buys. Average profit {: 5.2f}%. '
'Total profit was {: 7.3f}. Average duration {:5.1f} mins.').format(
len(results.index),
results.profit.mean() * 100.0,
results.profit.sum(),
results.duration.mean() * 5,
)
def buy_strategy_generator(params):
def populate_buy_trend(dataframe: DataFrame) -> DataFrame:
@@ -70,94 +165,14 @@ def buy_strategy_generator(params):
return populate_buy_trend
@pytest.mark.skipif(not os.environ.get('BACKTEST', False), reason="BACKTEST not set")
def test_hyperopt(backtest_conf, mocker):
mocked_buy_trend = mocker.patch('freqtrade.tests.test_backtesting.populate_buy_trend')
def start(args):
# TODO: parse args
backdata = load_backtesting_data()
processed = preprocess(backdata)
exchange._API = Bittrex({'key': '', 'secret': ''})
def optimizer(params):
mocked_buy_trend.side_effect = buy_strategy_generator(params)
results = backtest(backtest_conf, processed, mocker)
result = format_results(results)
total_profit = results.profit.sum() * 1000
trade_count = len(results.index)
trade_loss = 1 - 0.35 * exp(-(trade_count - TARGET_TRADES) ** 2 / 10 ** 5.2)
profit_loss = max(0, 1 - total_profit / 10000) # max profit 10000
# pylint: disable=W0603
global current_tries
current_tries += 1
print('{:5d}/{}: {}'.format(current_tries, TOTAL_TRIES, result))
return {
'loss': trade_loss + profit_loss,
'status': STATUS_OK,
'result': result
}
space = {
'mfi': hp.choice('mfi', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('mfi-value', 5, 25, 1)}
]),
'fastd': hp.choice('fastd', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('fastd-value', 10, 50, 1)}
]),
'adx': hp.choice('adx', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('adx-value', 15, 50, 1)}
]),
'rsi': hp.choice('rsi', [
{'enabled': False},
{'enabled': True, 'value': hp.quniform('rsi-value', 20, 40, 1)}
]),
'uptrend_long_ema': hp.choice('uptrend_long_ema', [
{'enabled': False},
{'enabled': True}
]),
'uptrend_short_ema': hp.choice('uptrend_short_ema', [
{'enabled': False},
{'enabled': True}
]),
'over_sar': hp.choice('over_sar', [
{'enabled': False},
{'enabled': True}
]),
'green_candle': hp.choice('green_candle', [
{'enabled': False},
{'enabled': True}
]),
'uptrend_sma': hp.choice('uptrend_sma', [
{'enabled': False},
{'enabled': True}
]),
'trigger': hp.choice('trigger', [
{'type': 'lower_bb'},
{'type': 'faststoch10'},
{'type': 'ao_cross_zero'},
{'type': 'ema5_cross_ema10'},
{'type': 'macd_cross_signal'},
{'type': 'sar_reversal'},
{'type': 'stochf_cross'},
{'type': 'ht_sine'},
]),
}
trials = Trials()
best = fmin(fn=optimizer, space=space, algo=tpe.suggest, max_evals=TOTAL_TRIES, trials=trials)
best = fmin(fn=optimizer, space=SPACE, algo=tpe.suggest, max_evals=TOTAL_TRIES, trials=trials)
print('\n\n\n\n==================== HYPEROPT BACKTESTING REPORT ==============================')
print('Best parameters {}'.format(best))
newlist = sorted(trials.results, key=itemgetter('loss'))
print('Result: {}'.format(newlist[0]['result']))
if __name__ == '__main__':
# for profiling with cProfile and line_profiler
pytest.main([__file__, '-s'])