freqtrade_origin/freqtrade/optimize/analysis/recursive.py

184 lines
7.3 KiB
Python
Raw Normal View History

2023-09-04 01:53:04 +00:00
import logging
import shutil
from copy import deepcopy
2023-09-12 10:54:25 +00:00
from datetime import timedelta
2023-09-04 01:53:04 +00:00
from pathlib import Path
2023-09-12 10:54:25 +00:00
from typing import Any, Dict, List
2023-09-04 01:53:04 +00:00
from pandas import DataFrame
from freqtrade.exchange import timeframe_to_minutes
from freqtrade.loggers.set_log_levels import (reduce_verbosity_for_bias_tester,
restore_verbosity_for_bias_tester)
from freqtrade.optimize.backtesting import Backtesting
2023-09-12 06:42:32 +00:00
from freqtrade.optimize.base_analysis import BaseAnalysis, VarHolder
2023-09-04 01:53:04 +00:00
logger = logging.getLogger(__name__)
2023-09-12 06:42:32 +00:00
class RecursiveAnalysis(BaseAnalysis):
2023-09-04 01:53:04 +00:00
def __init__(self, config: Dict[str, Any], strategy_obj: Dict):
2023-09-12 07:20:04 +00:00
self._startup_candle = config.get('startup_candle', [199, 399, 499, 999, 1999])
2023-09-12 06:42:32 +00:00
super().__init__(config, strategy_obj)
2023-09-12 10:50:39 +00:00
2023-09-04 02:45:25 +00:00
self.partial_varHolder_array: List[VarHolder] = []
self.partial_varHolder_lookahead_array: List[VarHolder] = []
2023-09-04 01:53:04 +00:00
2023-09-04 02:52:09 +00:00
self.dict_recursive: Dict[str, Any] = dict()
2023-09-04 01:53:04 +00:00
# For recursive bias check
# analyzes two data frames with processed indicators and shows differences between them.
def analyze_indicators(self):
2023-09-04 02:35:44 +00:00
2023-09-04 01:53:04 +00:00
pair_to_check = self.local_config['pairs'][0]
2023-09-04 02:35:44 +00:00
logger.info("Start checking for recursive bias")
2023-09-04 01:53:04 +00:00
# check and report signals
base_last_row = self.full_varHolder.indicators[pair_to_check].iloc[-1]
2023-09-04 02:38:13 +00:00
2023-09-04 01:53:04 +00:00
for part in self.partial_varHolder_array:
part_last_row = part.indicators[pair_to_check].iloc[-1]
compare_df = base_last_row.compare(part_last_row)
if compare_df.shape[0] > 0:
# print(compare_df)
for col_name, values in compare_df.items():
# print(col_name)
if 'other' == col_name:
continue
indicators = values.index
for indicator in indicators:
2023-09-04 02:45:25 +00:00
if (indicator not in self.dict_recursive):
2023-09-04 01:53:04 +00:00
self.dict_recursive[indicator] = {}
values_diff = compare_df.loc[indicator]
values_diff_self = values_diff.loc['self']
values_diff_other = values_diff.loc['other']
2023-09-04 02:35:44 +00:00
diff = (values_diff_other - values_diff_self) / values_diff_self * 100
2023-09-04 01:53:04 +00:00
2023-09-04 02:41:24 +00:00
self.dict_recursive[indicator][part.startup_candle] = f"{diff:.3f}%"
2023-09-04 01:53:04 +00:00
else:
logger.info("No variance on indicator(s) found due to recursive formula.")
2023-09-04 01:53:04 +00:00
break
# For lookahead bias check
# analyzes two data frames with processed indicators and shows differences between them.
def analyze_indicators_lookahead(self):
2023-09-04 02:35:44 +00:00
2023-09-04 01:53:04 +00:00
pair_to_check = self.local_config['pairs'][0]
2023-09-04 02:35:44 +00:00
logger.info("Start checking for lookahead bias on indicators only")
2023-09-04 01:53:04 +00:00
part = self.partial_varHolder_lookahead_array[0]
part_last_row = part.indicators[pair_to_check].iloc[-1]
date_to_check = part_last_row['date']
2023-09-04 02:35:44 +00:00
index_to_get = (self.full_varHolder.indicators[pair_to_check]['date'] == date_to_check)
base_row_check = self.full_varHolder.indicators[pair_to_check].loc[index_to_get].iloc[-1]
2023-09-04 01:53:04 +00:00
check_time = part.to_dt.strftime('%Y-%m-%dT%H:%M:%S')
logger.info(f"Check indicators at {check_time}")
# logger.info(f"vs {part_timerange} with {part.startup_candle} startup candle")
2023-09-04 02:35:44 +00:00
compare_df = base_row_check.compare(part_last_row)
2023-09-04 01:53:04 +00:00
if compare_df.shape[0] > 0:
# print(compare_df)
for col_name, values in compare_df.items():
# print(col_name)
if 'other' == col_name:
continue
indicators = values.index
for indicator in indicators:
logger.info(f"=> found lookahead in indicator {indicator}")
# logger.info("base value {:.5f}".format(values_diff_self))
# logger.info("part value {:.5f}".format(values_diff_other))
else:
logger.info("No lookahead bias on indicators found.")
2023-09-04 01:53:04 +00:00
2023-09-12 10:29:13 +00:00
def prepare_data(self, varholder: VarHolder, pairs_to_load: List[DataFrame]):
if 'freqai' in self.local_config and 'identifier' in self.local_config['freqai']:
# purge previous data if the freqai model is defined
# (to be sure nothing is carried over from older backtests)
path_to_current_identifier = (
Path(f"{self.local_config['user_data_dir']}/models/"
f"{self.local_config['freqai']['identifier']}").resolve())
# remove folder and its contents
if Path.exists(path_to_current_identifier):
shutil.rmtree(path_to_current_identifier)
prepare_data_config = deepcopy(self.local_config)
prepare_data_config['timerange'] = (str(self.dt_to_timestamp(varholder.from_dt)) + "-" +
str(self.dt_to_timestamp(varholder.to_dt)))
prepare_data_config['exchange']['pair_whitelist'] = pairs_to_load
backtesting = Backtesting(prepare_data_config, self.exchange)
self.exchange = backtesting.exchange
2023-09-12 10:29:13 +00:00
backtesting._set_strategy(backtesting.strategylist[0])
varholder.data, varholder.timerange = backtesting.load_bt_data()
backtesting.load_bt_data_detail()
varholder.timeframe = backtesting.timeframe
varholder.indicators = backtesting.strategy.advise_all_indicators(varholder.data)
2023-09-04 01:53:04 +00:00
def fill_partial_varholder(self, start_date, startup_candle):
2023-09-21 07:45:43 +00:00
logger.info(f"Calculating indicators using startup candle of {startup_candle}.")
2023-09-04 01:53:04 +00:00
partial_varHolder = VarHolder()
partial_varHolder.from_dt = start_date
partial_varHolder.to_dt = self.full_varHolder.to_dt
partial_varHolder.startup_candle = startup_candle
self.local_config['startup_candle_count'] = startup_candle
self.prepare_data(partial_varHolder, self.local_config['pairs'])
self.partial_varHolder_array.append(partial_varHolder)
def fill_partial_varholder_lookahead(self, end_date):
2023-09-21 07:45:43 +00:00
logger.info("Calculating indicators to test lookahead on indicators.")
2023-09-04 01:53:04 +00:00
partial_varHolder = VarHolder()
partial_varHolder.from_dt = self.full_varHolder.from_dt
partial_varHolder.to_dt = end_date
self.prepare_data(partial_varHolder, self.local_config['pairs'])
self.partial_varHolder_lookahead_array.append(partial_varHolder)
def start(self) -> None:
2023-09-12 06:42:32 +00:00
super().start()
2023-09-12 07:20:04 +00:00
reduce_verbosity_for_bias_tester()
2023-09-04 01:53:04 +00:00
start_date_full = self.full_varHolder.from_dt
end_date_full = self.full_varHolder.to_dt
timeframe_minutes = timeframe_to_minutes(self.full_varHolder.timeframe)
end_date_partial = start_date_full + timedelta(minutes=int(timeframe_minutes * 10))
self.fill_partial_varholder_lookahead(end_date_partial)
# restore_verbosity_for_bias_tester()
start_date_partial = end_date_full - timedelta(minutes=int(timeframe_minutes))
for startup_candle in self._startup_candle:
self.fill_partial_varholder(start_date_partial, int(startup_candle))
# Restore verbosity, so it's not too quiet for the next strategy
restore_verbosity_for_bias_tester()
self.analyze_indicators()
2023-09-04 02:35:44 +00:00
self.analyze_indicators_lookahead()