# --- Do not remove these libs --- from freqtrade.strategy.interface import IStrategy from typing import Dict, List from functools import reduce from pandas import DataFrame # -------------------------------- import talib.abstract as ta import freqtrade.vendor.qtpylib.indicators as qtpylib import numpy # noqa from freqtrade.strategy import CategoricalParameter, DecimalParameter, IntParameter import Config class SqueezeMomentum(IStrategy): """ Strategy based on LazyBear Squeeze Momentum Indicator (on TradingView.com) How to use it? > python3 ./freqtrade/main.py -s SqueezeMomentum """ # Hyperparameters # Buy hyperspace params: buy_params = { "buy_accel_enabled": True, "buy_adx": 14.0, "buy_adx_enabled": True, "buy_bb_enabled": False, "buy_bb_gain": 0.03, "buy_ema_enabled": True, "buy_macd_enabled": True, "buy_mfi": 6.0, "buy_mfi_enabled": False, "buy_period": 11, "buy_predict_enabled": True, "buy_sqz_band": 0.0112, } buy_period = IntParameter(3, 20, default=16, space="buy") buy_adx = DecimalParameter(1, 99, decimals=0, default=21, space="buy") buy_mfi = DecimalParameter(1, 30, decimals=0, default=14, space="buy") buy_sqz_band = DecimalParameter(0.003, 0.02, decimals=4, default=0.0165, space="buy") buy_bb_gain = DecimalParameter(0.01, 0.10, decimals=2, default=0.08, space="buy") buy_bb_enabled = CategoricalParameter([True, False], default=True, space="buy") # buy_sqz_enabled = CategoricalParameter([True, False], default=True, space="buy") buy_macd_enabled = CategoricalParameter([True, False], default=True, space="buy") buy_adx_enabled = CategoricalParameter([True, False], default=False, space="buy") buy_mfi_enabled = CategoricalParameter([True, False], default=True, space="buy") buy_ema_enabled = CategoricalParameter([True, False], default=True, space="buy") # buy_dc_enabled = CategoricalParameter([True, False], default=False, space="buy") buy_predict_enabled = CategoricalParameter([True, False], default=True, space="buy") buy_accel_enabled = CategoricalParameter([True, False], default=True, space="buy") sell_sar_enabled = CategoricalParameter([True, False], default=False, space="sell") sell_dc_enabled = CategoricalParameter([True, False], default=False, space="sell") sell_fisher = DecimalParameter(-1, 1, decimals=2, default=-0.30, space="sell") sell_standard_triggers = CategoricalParameter([True, False], default=False, space="sell") sell_hold = CategoricalParameter([True, False], default=True, space="sell") # set the startup candles count to the longest average used (EMA, EMA etc) startup_candle_count = max(buy_period.value, 20) # set common parameters minimal_roi = Config.minimal_roi trailing_stop = Config.trailing_stop trailing_stop_positive = Config.trailing_stop_positive trailing_stop_positive_offset = Config.trailing_stop_positive_offset trailing_only_offset_is_reached = Config.trailing_only_offset_is_reached stoploss = Config.stoploss timeframe = Config.timeframe process_only_new_candles = Config.process_only_new_candles use_exit_signal = Config.use_exit_signal exit_profit_only = Config.exit_profit_only ignore_roi_if_entry_signal = Config.ignore_roi_if_entry_signal order_types = Config.order_types def informative_pairs(self): """ Define additional, informative pair/interval combinations to be cached from the exchange. These pair/interval combinations are non-tradeable, unless they are part of the whitelist as well. For more information, please consult the documentation :return: List of tuples in the format (pair, interval) Sample: return [("ETH/USDT", "5m"), ("BTC/USDT", "15m"), ] """ return [] 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. """ # SMA - Simple Moving Average dataframe['sma'] = ta.SMA(dataframe, timeperiod=self.buy_period.value) dataframe['ema'] = ta.EMA(dataframe, timeperiod=self.buy_period.value) dataframe['tema'] = ta.TEMA(dataframe, timeperiod=self.buy_period.value) dataframe['ema3'] = ta.EMA(dataframe, timeperiod=3) dataframe['ema5'] = ta.EMA(dataframe, timeperiod=5) dataframe['ema10'] = ta.EMA(dataframe, timeperiod=10) # MACD macd = ta.MACD(dataframe) dataframe['macd'] = macd['macd'] dataframe['macdsignal'] = macd['macdsignal'] dataframe['macdhist'] = macd['macdhist'] # MFI dataframe['mfi'] = ta.MFI(dataframe) # ADX dataframe['adx'] = ta.ADX(dataframe) dataframe['dm_plus'] = ta.PLUS_DM(dataframe) dataframe['dm_minus'] = ta.MINUS_DM(dataframe) dataframe['dm_delta'] = dataframe['dm_plus'] - dataframe['dm_minus'] # SAR Parabolic dataframe['sar'] = ta.SAR(dataframe) # RSI dataframe['rsi'] = ta.RSI(dataframe) # Inverse Fisher transform on RSI, values [-1.0, 1.0] (https://goo.gl/2JGGoy) rsi = 0.1 * (dataframe['rsi'] - 50) dataframe['fisher_rsi'] = (numpy.exp(2 * rsi) - 1) / (numpy.exp(2 * rsi) + 1) # Bollinger Bands # bollinger = qtpylib.bollinger_bands(qtpylib.typical_price(dataframe), window=self.buy_period.value, stds=2) bollinger = qtpylib.weighted_bollinger_bands(qtpylib.typical_price(dataframe), window=20, stds=2) dataframe['bb_upperband'] = bollinger['upper'] dataframe['bb_mid'] = bollinger['mid'] dataframe['bb_lowerband'] = bollinger['lower'] dataframe["bb_gain"] = ((dataframe["bb_upperband"] - dataframe["close"]) / dataframe["close"]) # Keltner Channel keltner = qtpylib.keltner_channel(dataframe) dataframe["kc_upper"] = keltner["upper"] dataframe["kc_lower"] = keltner["lower"] dataframe["kc_middle"] = keltner["mid"] # Donchian Channels dataframe['dc_upper'] = ta.MAX(dataframe['high'], timeperiod=self.buy_period.value) dataframe['dc_lower'] = ta.MIN(dataframe['low'], timeperiod=self.buy_period.value) dataframe['dc_mid'] = ta.TEMA(((dataframe['dc_upper'] + dataframe['dc_lower']) / 2), timeperiod=self.buy_period.value) # Fibonacci Levels (of Donchian Channel) dataframe['dc_dist'] = (dataframe['dc_upper'] - dataframe['dc_lower']) dataframe['dc_hf'] = dataframe['dc_upper'] - dataframe['dc_dist'] * 0.236 # Highest Fib dataframe['dc_chf'] = dataframe['dc_upper'] - dataframe['dc_dist'] * 0.382 # Centre High Fib dataframe['dc_clf'] = dataframe['dc_upper'] - dataframe['dc_dist'] * 0.618 # Centre Low Fib dataframe['dc_lf'] = dataframe['dc_upper'] - dataframe['dc_dist'] * 0.764 # Low Fib # Squeeze Indicators. # 'on' means Bollinger Band lies completely within the Keltner Channel # 'off' means Keltner Channel lies completely within the Bollinger Band # Booleans are funky with dataframes, so just do an intermediate calculation dataframe['sqz_upper'] = (dataframe['bb_upperband'] - dataframe["kc_upper"]) dataframe['sqz_lower'] = (dataframe['bb_lowerband'] - dataframe["kc_lower"]) dataframe['sqz_on'] = ((dataframe['sqz_upper'] < 0) & (dataframe['sqz_lower'] > 0)) dataframe['sqz_off'] = ((dataframe['sqz_upper'] > 0) & (dataframe['sqz_lower'] < 0)) # Momentum # value is: Close - Moving Average( (Donchian midline + EMA) / 2 ) # get momentum value by running linear regression on delta dataframe['sqz_ave'] = ta.TEMA(((dataframe['dc_mid'] + dataframe['tema']) / 2), timeperiod=self.buy_period.value) dataframe['sqz_delta'] = ta.TEMA((dataframe['close'] - dataframe['sqz_ave']), timeperiod=30) # timeperiod = self.buy_period.value) dataframe['sqz_val'] = ta.LINEARREG(dataframe['sqz_delta'], timeperiod=self.buy_period.value) # the angle will show turnaround points (at 0). Use just a little averaging to avoid 'wiggles' #dataframe['sqz_angle'] = ta.LINEARREG_ANGLE(dataframe['sqz_delta'], timeperiod=3) dataframe['sqz_angle'] = ta.LINEARREG_SLOPE(dataframe['sqz_delta'], timeperiod=3) dataframe['sqz_a'] = ta.LINEARREG(dataframe['sqz_angle'], timeperiod=3) dataframe['sqz_accel'] = ta.LINEARREG_SLOPE(dataframe['sqz_a'], timeperiod=3) return dataframe def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: conditions = [] # GUARDS AND TRENDS # check that volume is not 0 (can happen in testing, or if there are issues with exchange data) # conditions.append(dataframe['volume'] > 0) # ADX with DM+ > DM- indicates uptrend if self.buy_adx_enabled.value: conditions.append( # (dataframe['adx'] > self.buy_adx.value) (dataframe['adx'] >= self.buy_adx.value) & (dataframe['dm_plus'] >= dataframe['dm_minus']) ) if self.buy_mfi_enabled.value: conditions.append(dataframe['mfi'] <= self.buy_mfi.value) # only buy if close is below EMA if self.buy_ema_enabled.value: conditions.append(dataframe['close'] < dataframe['ema']) # only buy if close above High Fibonacci (i.e. potential breakout) # if self.buy_dc_enabled.value: # conditions.append( # (dataframe['close'] >= dataframe['dc_hf']) | # (dataframe['open'] >= dataframe['dc_hf']) # ) # MACD -ve (this is the opposite of common sense, because it is a lagging indicator) if self.buy_macd_enabled.value: conditions.append(dataframe['macd'] <= 0) conditions.append(dataframe['macd'] <= dataframe['macdsignal']) # potential gain > goal if self.buy_bb_enabled.value: conditions.append(dataframe['bb_gain'] >= self.buy_bb_gain.value) # squeeze is 'on' # if self.buy_sqz_enabled.value: # conditions.append(dataframe['sqz_on']) # We can (try to) predict an upcoming swing (up) by looking for a reversal during an 'off' period if self.buy_predict_enabled.value: # check for startup issue conditions.append( (dataframe['sqz_val'].notnull()) & (dataframe['sqz_val'].shift(5).notnull()) ) # Green candle conditions.append(dataframe['close'] > dataframe['open']) # TRIGGERS # squeeze values are -ve but turning around conditions.append(dataframe['sqz_val'] < -self.buy_sqz_band.value) if self.buy_accel_enabled.value: conditions.append(qtpylib.crossed_above(dataframe['sqz_accel'], 0)) else: conditions.append(qtpylib.crossed_above(dataframe['sqz_angle'], 0)) else: # during back testing, data can be undefined, so check # conditions.append(dataframe['sqz_upper'].notnull()) # TRIGGERS # Momentum goes positive , and is increasing conditions.append(qtpylib.crossed_above(dataframe['sqz_val'], 0)) # current candle is green # conditions.append(dataframe['close'] > dataframe['open']) # build the dataframe using the conditions if conditions: dataframe.loc[ reduce(lambda x, y: x & y, conditions), 'buy'] = 1 return dataframe def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: """ Based on TA indicators, populates the sell signal for the given dataframe :param dataframe: DataFrame :return: DataFrame with buy column """ # if hold flag is set then don't issue any sell signals at all (rely on ROI and stoploss) if self.sell_hold.value: dataframe.loc[ (dataframe['close'].notnull()), 'sell'] = 0 return dataframe conditions = [] # GUARDS AND TRENDS # check that volume is not 0 (can happen in testing, or if there are issues with exchange data) #conditions.append(dataframe['volume'] > 0) # only sell if close is above SAR if self.sell_sar_enabled.value: conditions.append(dataframe['close'] > dataframe['sar']) # only buy if close below Low Fibonacci (i.e. potential breakdown) if self.sell_dc_enabled.value: conditions.append(dataframe['close'] <= dataframe['dc_lf']) # squeeze is 'off' # if self.buy_sqz_enabled.value: # conditions.append(dataframe['sqz_on'] != True) # We can (try to) predict an upcoming swing (down) by looking for a reversal during an 'on' period if self.buy_predict_enabled: # TRIGGERS # squeeze values are +ve but turning around conditions.append(dataframe['sqz_val'] > self.buy_sqz_band.value) #conditions.append(qtpylib.crossed_below(dataframe['sqz_angle'], 0)) conditions.append(qtpylib.crossed_below(dataframe['sqz_accel'], 0)) # conditions.append( # (dataframe['sqz_val'].shift(1) > self.buy_sqz_band.value) & # (dataframe['sqz_val'] < dataframe['sqz_val'].shift(1)) & # (dataframe['sqz_val'].shift(1) >= dataframe['sqz_val'].shift(2)) & # (dataframe['sqz_val'].shift(2) >= dataframe['sqz_val'].shift(3)) & # (dataframe['sqz_val'].shift(3) >= dataframe['sqz_val'].shift(4)) # ) else: # during back testing, data can be undefined, so check conditions.append(dataframe['sqz_val'].notnull()) # TRIGGERS # Momentum goes negative conditions.append(qtpylib.crossed_below(dataframe['sqz_val'], 0)) #conditions.append( # (dataframe['sqz_val'] < 0) & # (dataframe['sqz_val'].shift(1) < 0) & # (dataframe['sqz_val'].shift(2) >= 0) #) # 'standard' sell triggers orconditions = [] if self.sell_standard_triggers.value: orconditions.append( (dataframe['fisher_rsi'] > self.sell_fisher.value) & (dataframe['sar'] > dataframe['close']) ) # build the dataframe using the conditions r1 = False r2 = False if conditions: r1 = reduce(lambda x, y: x & y, conditions) if orconditions: r2 = reduce(lambda x, y: x & y, orconditions) dataframe.loc[(r1 | r2), 'sell'] = 1 return dataframe # +--------+---------+----------+--------------------------+--------------+-------------------------------+-----------------+-------------+-------------------------------+ # | Best | Epoch | Trades | Win Draw Loss Win% | Avg profit | Profit | Avg duration | Objective | Max Drawdown (Acct) | # |--------+---------+----------+--------------------------+--------------+-------------------------------+-----------------+-------------+-------------------------------| # | * Best | 1/50 | 11 | 7 0 4 63.6 | 0.79% | 895.843 USDT (8.96%) | 0 days 19:11:00 | -895.843 | 0.808 USDT (0.01%) | # | * Best | 4/50 | 32 | 18 3 11 56.2 | 0.67% | 2339.471 USDT (23.39%) | 0 days 19:20:00 | -2,339.47095 | 209.676 USDT (1.67%) | # | Best | 38/50 | 44 | 27 2 15 61.4 | 0.50% | 2412.368 USDT (24.12%) | 0 days 15:02:00 | -2,412.36823 | 157.722 USDT (1.25%) | # | Best | 45/50 | 34 | 21 4 9 61.8 | 0.67% | 2516.440 USDT (25.16%) | 0 days 17:34:00 | -2,516.43966 | 22.785 USDT (0.19%) |