# pragma pylint: disable=missing-docstring, invalid-name, pointless-string-statement import logging # add trading_environments to system path import sys from collections import defaultdict from datetime import datetime from functools import partial from pathlib import Path from typing import Dict, Optional import freqtrade.vendor.qtpylib.indicators as qtpylib import numpy as np # noqa import pandas as pd # noqa import pandas_ta as pta import talib.abstract as ta from freqtrade.exchange import timeframe_to_prev_date from freqtrade.persistence import Trade from freqtrade.strategy import informative from freqtrade.strategy.interface import IStrategy from lazyft.space_handler import SpaceHandler from lazyft.strategy import load_strategy from pandas import DataFrame from sb3_contrib import RecurrentPPO from sklearn.preprocessing import robust_scale from stable_baselines3 import A2C from stable_baselines3.ppo.ppo import PPO # from trading_environments.my_freqtrade_env import Actions logger = logging.getLogger(__name__) COLUMNS_FILTER = [ "date", "open", "close", "high", "low", "buy", "sell", "volume", "buy_tag", "exit_tag", ] from scipy import stats stats.zscore = partial(stats.zscore, nan_policy="omit") class SagesGym5(IStrategy): sh = SpaceHandler(__file__, disable=__name__ != __qualname__) model_path = sh.get_setting("model_path") # # If you've used SimpleROIEnv then use this minimal_roi # minimal_roi = { # "720": -10, # "600": 0.00001, # "60": 0.01, # "30": 0.02, # "0": 0.03 # } minimal_roi = {"0": 100} stoploss = -0.99 # Trailing stop: trailing_stop = False trailing_stop_positive = 0.001 trailing_stop_positive_offset = 0.017 trailing_only_offset_is_reached = True timeframe = "15m" use_sell_signal = True # Run "populate_indicators()" only for new candle. process_only_new_candles = False startup_candle_count: int = 10 model = None window_size = 10 timeperiods = [7, 14, 21] percent_of_balance_dict = {} def __init__(self, config: dict) -> None: super().__init__(config) # self.custom_strategy = load_strategy('BatsContest', self.config) self.model = None try: # get the file that starts with "best_model_" in the models/ directory # list files in the directory # files = Path('models/').glob('final_model_*') # get the first file # model_file = next(files) model_file = Path( "models/best_model_SagesGym5_SagesFreqtradeEnv_RecurrentPPO_20220501_090806.zip" ) # assert model_file.exists(), f'Model file "{model_file}" does not exist.' self.model = RecurrentPPO.load( str(model_file) ) # Note: Make sure you use the same policy as the one used to train self.window_size = self.model.observation_space.shape[0] print("Observation space/Window size", self.window_size) except Exception as e: logger.exception(f"Could not load model: {e}") raise else: logger.info(f"Loaded model: {model_file}") # @informative('4h', 'BTC/{stake}') # @informative('2h', 'BTC/{stake}') # @informative('1h', 'BTC/{stake}') def populate_indicators_btc_4h( self, dataframe: DataFrame, metadata: dict ) -> DataFrame: dataframe[f"rsi"] = stats.zscore(ta.RSI(dataframe["close"], timeperiod=30)) dataframe["top"] = stats.zscore(dataframe["close"].rolling(window=30).max()) return dataframe def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame: """ Adds several different TA indicators to the given DataFrame Performance Note: For the best performance be frugal on the number of indicators you are using. Let uncomment only the indicator you are using in your strategies or your hyperopt configuration, otherwise you will waste your memory and CPU usage. :param dataframe: Raw data from the exchange and parsed by parse_ticker_dataframe() :param metadata: Additional information, like the currently traded pair :return: a Dataframe with all mandatory indicators for the strategies """ logger.info(f'Calculating TA indicators for {metadata["pair"]}') # indicators = dataframe[dataframe.columns[~dataframe.columns.isin(COLUMNS_FILTER)]] # assert all(indicators.max() < 1.00001) and all( # indicators.min() > -0.00001 # ), "Error, values are not normalized!" # logger.info(f'{metadata["pair"]} - indicators populated!') # dataframe = self.custom_strategy.populate_indicators(dataframe, metadata) # rsi dataframe["return"] = dataframe["close"].pct_change(1) return dataframe def populate_buy_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: """ Based on TA indicators, populates the buy signal for the given dataframe :param dataframe: DataFrame populated with indicators :param metadata: Additional information, like the currently traded pair :return: DataFrame with buy column """ # dataframe['buy'] = self.rl_model_predict(dataframe) assert self.model is not None, "Model is not loaded." logger.info(f'Populating buy signal for {metadata["pair"]}') dataframe = self.rl_model_predict(dataframe, metadata["pair"]) # dataframe["buy"] = (action == 1).astype("int") logger.info(f'{metadata["pair"]} - buy signal populated!') self.wallets.get_all_balances return dataframe def populate_sell_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: """ Based on TA indicators, populates the sell signal for the given dataframe :param dataframe: DataFrame populated with indicators :param metadata: Additional information, like the currently traded pair :return: DataFrame with buy column """ logger.info(f'Populating sell signal for {metadata["pair"]}') # action = self.rl_model_predict(dataframe, metadata["pair"]) # dataframe["sell"] = (action == 2).astype("int") # print number of sell signals print(dataframe["sell"].value_counts(), "sell signals") logger.info(f'{metadata["pair"]} - sell signal populated!') return dataframe def rl_model_predict(self, dataframe: DataFrame, pair: str): action_output = pd.DataFrame(np.zeros((len(dataframe), 1))) # multiplier_output = pd.DataFrame(np.zeros((len(dataframe), 1))) indicators = dataframe[["return", "volume"]] indicators = pd.DataFrame( robust_scale(indicators.values), columns=indicators.columns, index=indicators.index, ) print(indicators.describe()) to_numpy = indicators.fillna(0).to_numpy() indicators["date"] = dataframe["date"] # start index where all indicators are available # print(f'{indicators.shape}') # TODO: This is slow and ugly, must use .rolling # logger.info("Indicators head: %s", indicators[:5]) for window in range(self.window_size, len(dataframe)): start = window - self.window_size end = window observation = to_numpy[start:end] # print("observation", observation) res, _ = self.model.predict(observation, deterministic=True) action = res if action == 1: logger.info("Observation: %s", indicators[start:end]) action_output.loc[end] = action dataframe["buy"] = (action_output == 1).astype("int") dataframe["sell"] = (action_output == 2).astype("int") return dataframe # def custom_stake_amount( # self, # pair: str, # current_time: datetime, # current_rate: float, # proposed_stake: float, # min_stake: float, # max_stake: float, # entry_tag: Optional[str], # **kwargs, # ) -> float: # dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe) # last_candle = dataframe.iloc[-1].squeeze() # pob = self.percent_of_balance_dict[last_candle.name.astype("int")] # if pob > 0: # return pob / 10 * self.wallets.get_available_stake_amount() def normalize(data, min_value, max_value): return (data - min_value) / (max_value - min_value) def rolling_z_score(data: pd.Series, window): # use stats.zscore to calculate rolling z-score r = data.rolling(window=window) m = r.mean().shift(1) s = r.std(ddof=0).shift(1) z = (data - m) / s return z