import logging import os import sys import json from functools import reduce import numpy import numpy as np # noqa import pandas as pd # noqa from pandas import DataFrame from typing import Optional, Union, List from pandas_ta import stdev from freqtrade.enums import ExitCheckTuple from freqtrade.persistence import Trade, Order from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, IStrategy, IntParameter, informative) from datetime import timedelta, datetime, timezone import talib.abstract as ta import freqtrade.vendor.qtpylib.indicators as qtpylib import pandas_ta as pta """ Author of this strategy is Milan Kajnar (mkajnar@mikesoft.cz) """ """ Do not publish my strategy to another sites without my name please! """ class HPStrategyV8UltraDCACSL(IStrategy): INTERFACE_VERSION = 3 timeframe = '5m' leverage_value = 3 stoploss = -0.3 csl = {} minimal_roi = { "0": 0.015 * leverage_value } allprofits = {} candle_open_prices = {} max_open_trades = 10 process_only_new_candles = True startup_candle_count = 50 trailing_stop = True trailing_only_offset_is_reached = True trailing_stop_positive = 0.006 * leverage_value trailing_stop_positive_offset = 0.015 * leverage_value use_exit_signal = True use_custom_stoploss = True ignore_roi_if_entry_signal = True position_adjustment_enable = True custom_tp_pct = DecimalParameter(0.005, 0.50, default=0.005, decimals=3, space='sell', optimize=False) dca_threshold_pct_k = DecimalParameter(0.90, 0.99, default=0.94, decimals=2, space='buy', optimize=position_adjustment_enable) dca_threshold_pct = DecimalParameter(0.1, 0.5, default=0.05, decimals=2, space='buy', optimize=position_adjustment_enable) stoploss = (dca_threshold_pct.value - 1.25) * leverage_value dca_multiplier = DecimalParameter(1.2, 10, default=3, decimals=2, space='buy', optimize=position_adjustment_enable) dca_limit = IntParameter(1, 5, default=1, space='buy', optimize=position_adjustment_enable) donchian_period = IntParameter(5, 50, default=23, space='buy', optimize=False) rsi_treshold = IntParameter(10, 50, default=35, space='buy', optimize=False) cci_treshold = IntParameter(-200, -10, default=-100, space='buy', optimize=False) kick_off_threshold = DecimalParameter(-0.99, 0, default=-0.32, decimals=2, space='sell', optimize=False) pct3_buy_threshold = DecimalParameter(-0.999, -0.005, default=-0.01, decimals=3, space='buy', optimize=False) red_candles_before_buy = IntParameter(1, 5, default=1, space='buy', optimize=False) candle_time_threshold = DecimalParameter(0.01, 1, default=0.36, decimals=2, space='buy', optimize=False) trade_timeout = IntParameter(1, 48, default=12, space='sell', optimize=False) max_tradable_ratio = DecimalParameter(0.01, 1, default=0.75, space='buy', optimize=False) max_trades = IntParameter(1, 30, default=max_open_trades, space='buy', optimize=False) price_change_1 = DecimalParameter(0.001, 0.050, default=0.003, decimals=3, space='buy', optimize=True) price_change_2 = DecimalParameter(0.005, 0.150, default=0.025, decimals=3, space='buy', optimize=True) exit_profit_offset = 0.001 exit_profit_only = False order_types = { 'entry': 'market', 'exit': 'market', 'stoploss': 'market', 'stoploss_on_exchange': False } def calc_donchian_channels(self, dataframe, period: int): dataframe["upperDon"] = dataframe["high"].rolling(period).max() dataframe["lowerDon"] = dataframe["low"].rolling(period).min() dataframe["midDon"] = (dataframe["upperDon"] + dataframe["lowerDon"]) / 2 return dataframe def mid_don_cross_over(self, dataframe): dataframe["position_m"] = np.nan dataframe["position_m"] = np.where(dataframe["close"] > dataframe["midDon"], 1, dataframe["position_m"]) dataframe["position_m"] = dataframe["position_m"].ffill().fillna(0) return dataframe def don_channel_breakout(self, dataframe): dataframe["position_b"] = np.nan dataframe["position_b"] = np.where(dataframe["close"] > dataframe["upperDon"].shift(1), 1, dataframe["position_b"]) dataframe["position_b"] = dataframe["position_b"].ffill().fillna(0) return dataframe def don_reversal(self, dataframe): dataframe["position_r"] = np.nan dataframe["position_r"] = np.where(dataframe["close"] < dataframe["lowerDon"].shift(1), 1, dataframe["position_r"]) dataframe["position_r"] = dataframe["position_r"].ffill().fillna(0) return dataframe def leverage(self, pair: str, current_time: datetime, current_rate: float, proposed_leverage: float, max_leverage: float, entry_tag: Optional[str], side: str, **kwargs) -> float: return self.leverage_value def calculate_price_change_coefficient(self, dataframe: DataFrame, candles_count: int = 1) -> DataFrame: dataframe['price_change_pct'] = dataframe['close'].pct_change(periods=candles_count) dataframe['price_change_coeff'] = dataframe['price_change_pct'].apply( lambda x: max(min(x, 0.99), -0.99) if x != 0 else 0.01 * np.sign(x) ) dataframe[f'price_change_coeff_{candles_count}'] = dataframe['price_change_coeff'].rolling( window=candles_count).mean() * self.leverage_value return dataframe def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame: dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14) dataframe['cci'] = ta.CCI(dataframe, timeperiod=20) dataframe = self.calc_swings(dataframe) dataframe = self.calc_donchian_channels(dataframe=dataframe, period=self.donchian_period.value) dataframe = self.mid_don_cross_over(dataframe=dataframe) dataframe = self.don_reversal(dataframe=dataframe) dataframe = self.don_channel_breakout(dataframe=dataframe) dataframe = self.calculate_price_change_coefficient(dataframe, 3) dataframe['price_change'] = dataframe['close'].pct_change() * 100 significant_loss_threshold = -5 # e.g., -5% change recovery_threshold = 3 # e.g., 3% increase from previous low dataframe['significant_loss'] = dataframe['price_change'] < significant_loss_threshold dataframe['recovery'] = ((dataframe['significant_loss'].shift(1) == True) & (dataframe['price_change'] > recovery_threshold)) dataframe['loss_and_recovery_signal'] = ((dataframe['significant_loss'] == True) & (dataframe['recovery'].shift(-1) == True)).astype(int) return dataframe def calc_swings(self, dataframe): dataframe['swing_low'] = (dataframe['close'].shift(2) > dataframe['close'].shift(1)) & \ (dataframe['close'].shift(1) < dataframe['close']).astype(int) dataframe['swing_high'] = (dataframe['close'].shift(2) < dataframe['close'].shift(1)) & \ (dataframe['close'].shift(1) > dataframe['close']).astype(int) return dataframe def check_red_candles(self, dataframe: DataFrame, n: int) -> DataFrame: red_candle = dataframe['close'] < dataframe['open'] dataframe.loc[:, 'red_candles_in_row'] = red_candle.rolling(window=n).sum() == n return dataframe def timeframe_to_minutes(self, timeframe: str) -> int: if 'm' in timeframe: return int(timeframe.replace('m', '')) elif 'h' in timeframe: return int(timeframe.replace('h', '')) * 60 elif 'd' in timeframe: return int(timeframe.replace('d', '')) * 1440 else: raise ValueError(f"Unsupported timeframe: {timeframe}") def is_candle_open_more_than_threshold(self, dataframe: DataFrame, threshold: float) -> DataFrame: candle_timeframe_minutes = self.timeframe_to_minutes(timeframe=self.timeframe) t_time = pd.to_timedelta(candle_timeframe_minutes * threshold, unit='minutes') current_time = pd.Timestamp.utcnow() dataframe.loc[:, 'time_since_open'] = current_time - dataframe['date'] dataframe.loc[:, 'is_open_more_than_threshold'] = dataframe['time_since_open'] >= t_time return dataframe def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: self.check_red_candles(dataframe, self.red_candles_before_buy.value) self.is_candle_open_more_than_threshold(dataframe, threshold=self.candle_time_threshold.value) dataframe.loc[ ( ( (dataframe['position_r'] == 1) | (dataframe['swing_low'] == 1) | (dataframe['position_m'] == 1) | (dataframe['position_b'] == 1) ) & ( (dataframe['red_candles_in_row']) & (dataframe['is_open_more_than_threshold']) & (dataframe['rsi'] < self.rsi_treshold.value) & (dataframe['cci'] < self.cci_treshold.value) ) & ( ( (dataframe['price_change_coeff_3'] < -self.price_change_1.value) & (dataframe['price_change_coeff_3'] > -self.price_change_2.value) ) | ( (dataframe['price_change_coeff_3'] > self.price_change_1.value) & (dataframe['price_change_coeff_3'] < self.price_change_2.value) ) ) ), ['enter_long', 'enter_tag'] ] = (1, 'swing_low') return dataframe def custom_exit(self, pair: str, trade: Trade, current_time: datetime, current_rate: float, current_profit: float, **kwargs) -> Optional[Union[str, bool]]: dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe) last_candle = dataframe.iloc[-1].squeeze() if last_candle['price_change_coeff_3'] <= self.kick_off_threshold.value: return 'kick_off' def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: dataframe.loc[ ( (dataframe['swing_high'] == 1) ), ['exit_long', 'exit_tag'] ] = (1, 'swing_high') dataframe.loc[ ( (dataframe['price_change_coeff_3'] <= self.kick_off_threshold.value) ), ['exit_long', 'exit_tag'] ] = (1, 'kick_off') return dataframe def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float, rate: float, time_in_force: str, exit_reason: str, current_time: datetime, **kwargs) -> bool: profit_ratio = trade.calc_profit_ratio(rate) if 'swing' in exit_reason or 'trailing' in exit_reason: return profit_ratio > self.exit_profit_offset return True def custom_stoploss(self, pair: str, trade: 'Trade', current_time: datetime, current_rate: float, current_profit: float, **kwargs) -> float: stoploss_value = json.loads(self.csl.get(pair, '{}')).get('sl', self.stoploss) return stoploss_value def custom_stake_amount(self, pair: str, current_time: datetime, current_rate: float, proposed_stake: float, min_stake: Optional[float], max_stake: float, leverage: float, entry_tag: Optional[str], side: str, **kwargs) -> float: return min(max_stake, (max_stake * self.max_tradable_ratio.value) / (self.max_trades.value * 1.5)) def adjust_trade_position(self, trade: Trade, current_time: datetime, current_rate: float, current_profit: float, min_stake: Optional[float], max_stake: float, current_entry_rate: float, current_exit_rate: float, current_entry_profit: float, current_exit_profit: float, **kwargs) -> Optional[float]: dataframe, _ = self.dp.get_analyzed_dataframe(trade.pair, self.timeframe) if len(dataframe) < 3: return None last_candle, second_last_candle, third_last_candle = dataframe.iloc[-1], dataframe.iloc[-2], dataframe.iloc[-3] candle_open_price = last_candle['open'] if self.candle_open_prices.get(trade.pair) == candle_open_price: return None self.candle_open_prices[trade.pair] = candle_open_price is_last_candle_green = last_candle['close'] > last_candle['open'] is_previous_candles_red = all( candle['close'] < candle['open'] for candle in [second_last_candle, third_last_candle]) dynamic_threshold = -self.dca_threshold_pct.value * self.leverage_value if current_profit <= dynamic_threshold and is_last_candle_green and is_previous_candles_red: logging.info(f'{current_time} Adjusting position for {trade.pair} with {trade.stake_amount}. ' f'Last two candles were red, and current candle is green.') return min(max_stake, trade.stake_amount * self.dca_multiplier.value) return None