# source: https://raw.githubusercontent.com/mkajnar/WolfStrategy/279bf0e8aea4b2df801d1e03eaae5e6aa8fcfa92/WolfStrategyV1.py """ Github_mkajnar_WolfStrategy__WolfStrategyV1__20260212_003143 - Long -> Short Profit Flip Strategy ===================================================== Donchian Channel bounce Long with DCA + 100x Short on breakout down. Phases: 1. Long: Enter on Donchian lower bounce + RSI oversold + trend filter + bullish reversal + volume spike + bounce confirmation - DCA with cooling period, progressive spacing, support detection - Exit via ATR-based stepped trailing stop or ROI 2. Short: Enter on Donchian breakout down after profitable Long exit - MACD histogram + volume breakdown + EMA confirmation - 100x leverage, strict TP - Funded from profit bucket (Long phase profit) Version: 3.0.0 """ import datetime import logging from typing import Optional, Union, List, Tuple, Dict import numpy as np import talib.abstract as ta from pandas import DataFrame from freqtrade.persistence import Trade from freqtrade.strategy import ( DecimalParameter, IStrategy, IntParameter ) logger = logging.getLogger(__name__) class Github_mkajnar_WolfStrategy__WolfStrategyV1__20260212_003143(IStrategy): INTERFACE_VERSION = 3 can_short = True timeframe = '15m' # -- Only 1 open trade at a time -- max_open_trades = 1 # -- Minimal ROI: only for Long -- minimal_roi = { "0": 0.06, # 6% default ROI "60": 0.04, # 4% after 60 min "180": 0.025, # 2.5% after 3h "720": 0.01 # 1% after 12h } # -- Stoploss: effectively disabled for Long, DCA handles drawdowns -- stoploss = -0.99 # -- Trailing stop: will be overridden by custom_stoploss -- trailing_stop = False # -- Order types -- order_types = { 'entry': 'market', 'exit': 'market', 'stoploss': 'market', 'stoploss_on_exchange': True } # -- Signals -- use_exit_signal = True exit_profit_only = False # -- DCA -- position_adjustment_enable = True # ================================================================ # HYPEROPT PARAMETERS - BUY SPACE # ================================================================ # Entry signal tuning (3 params optimized) donchian_period = IntParameter(14, 30, default=20, space='buy', optimize=True) rsi_oversold = IntParameter(25, 45, default=35, space='buy', optimize=True) rsi_period = IntParameter(10, 18, default=14, space='buy', optimize=False) atr_period = IntParameter(10, 20, default=14, space='buy', optimize=False) # EMA: fixed - these are standard values, no need to optimize ema_fast = IntParameter(20, 100, default=50, space='buy', optimize=False) ema_slow = IntParameter(100, 300, default=200, space='buy', optimize=False) # DCA parameters (3 params optimized) dca_max_count = IntParameter(2, 6, default=4, space='buy', optimize=True) dca_base_drop = DecimalParameter(0.01, 0.04, default=0.02, space='buy', optimize=True) dca_multiplier = DecimalParameter(1.2, 2.0, default=1.5, space='buy', optimize=True) dca_cooling_candles = IntParameter(2, 8, default=3, space='buy', optimize=False) # Long leverage: fixed long_leverage_min = DecimalParameter(2.0, 4.0, default=2.0, space='buy', optimize=False) long_leverage_max = DecimalParameter(5.0, 10.0, default=7.0, space='buy', optimize=False) # ================================================================ # HYPEROPT PARAMETERS - SELL SPACE # ================================================================ # Short exit (1 param) short_take_profit = DecimalParameter(0.015, 0.05, default=0.025, space='sell', optimize=True) # Trailing stop parameters (2 params) trailing_atr_mult = DecimalParameter(1.0, 2.5, default=1.5, space='sell', optimize=True) breakeven_trigger = DecimalParameter(0.008, 0.025, default=0.012, space='sell', optimize=True) # Partial profit taking (1 param) partial_profit_pct = DecimalParameter(0.015, 0.04, default=0.025, space='sell', optimize=True) # Short max stake cap (USDT) - fixed short_max_stake = DecimalParameter(20.0, 100.0, default=50.0, space='sell', optimize=False) # ================================================================ # RUNTIME STATE # ================================================================ custom_profit_bucket: Dict[str, float] = {} awaiting_short: Dict[str, bool] = {} last_buy_price: Dict[str, float] = {} dca_count: Dict[str, int] = {} last_dca_candle: Dict[str, int] = {} # candle index of last DCA partial_profit_taken: Dict[str, bool] = {} # track partial TP loss_cooldown_until: Dict[str, datetime.datetime] = {} # cooldown after loss # ================================================================ # INDICATORS # ================================================================ def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame: pair = metadata['pair'] # -- Donchian Channels -- period = self.donchian_period.value dataframe['dc_upper'] = dataframe['high'].rolling(window=period).max() dataframe['dc_lower'] = dataframe['low'].rolling(window=period).min() dataframe['dc_mid'] = (dataframe['dc_upper'] + dataframe['dc_lower']) / 2 # -- RSI -- dataframe['rsi'] = ta.RSI(dataframe, timeperiod=self.rsi_period.value) # -- ATR (for leverage + trailing) -- dataframe['atr'] = ta.ATR(dataframe, timeperiod=self.atr_period.value) dataframe['atr_pct'] = dataframe['atr'] / dataframe['close'] * 100 # -- EMA Trend Filter -- dataframe['ema_fast'] = ta.EMA(dataframe, timeperiod=self.ema_fast.value) dataframe['ema_slow'] = ta.EMA(dataframe, timeperiod=self.ema_slow.value) # -- MACD -- macd = ta.MACD(dataframe, fastperiod=12, slowperiod=26, signalperiod=9) dataframe['macd'] = macd['macd'] dataframe['macd_signal'] = macd['macdsignal'] dataframe['macd_hist'] = macd['macdhist'] # -- Bollinger Bands -- bollinger = ta.BBANDS(dataframe, timeperiod=20, nbdevup=2.0, nbdevdn=2.0) dataframe['bb_upper'] = bollinger['upperband'] dataframe['bb_lower'] = bollinger['lowerband'] dataframe['bb_mid'] = bollinger['middleband'] dataframe['bb_pctb'] = (dataframe['close'] - dataframe['bb_lower']) / \ (dataframe['bb_upper'] - dataframe['bb_lower']).replace(0, np.nan) # -- Volume analysis -- dataframe['volume_sma'] = dataframe['volume'].rolling(window=20).mean() dataframe['volume_ratio'] = dataframe['volume'] / dataframe['volume_sma'].replace(0, np.nan) # -- Support detection: pivot lows (fractal-like) -- dataframe['pivot_low'] = ( (dataframe['low'] < dataframe['low'].shift(1)) & (dataframe['low'] < dataframe['low'].shift(2)) & (dataframe['low'] < dataframe['low'].shift(-1)) & (dataframe['low'] < dataframe['low'].shift(-2)) ) dataframe['support'] = dataframe['low'].where(dataframe['pivot_low']).ffill() dataframe['support'] = dataframe['support'].rolling(window=50, min_periods=1).min() # -- Donchian position: 0 = at lower, 1 = at upper -- dc_range = dataframe['dc_upper'] - dataframe['dc_lower'] dataframe['dc_position'] = ( (dataframe['close'] - dataframe['dc_lower']) / dc_range.replace(0, np.nan) ) # -- Green / Red candles -- dataframe['green_candle'] = dataframe['close'] > dataframe['open'] dataframe['red_candle'] = dataframe['close'] < dataframe['open'] # -- RSI direction -- dataframe['rsi_rising'] = dataframe['rsi'] > dataframe['rsi'].shift(1) dataframe['rsi_falling'] = dataframe['rsi'] < dataframe['rsi'].shift(1) # -- MACD crossover -- dataframe['macd_cross_above'] = ( (dataframe['macd'] > dataframe['macd_signal']) & (dataframe['macd'].shift(1) <= dataframe['macd_signal'].shift(1)) ) dataframe['macd_cross_below'] = ( (dataframe['macd'] < dataframe['macd_signal']) & (dataframe['macd'].shift(1) >= dataframe['macd_signal'].shift(1)) ) # -- Donchian breakout down: close below previous lower -- dataframe['dc_breakout_down'] = ( dataframe['close'] < dataframe['dc_lower'].shift(1) ) # -- Volume declining (for DCA - selling exhaustion) -- dataframe['volume_declining'] = ( (dataframe['volume'] < dataframe['volume'].shift(1)) & (dataframe['volume'].shift(1) < dataframe['volume'].shift(2)) ) # -- Candle index for cooling period tracking -- dataframe['candle_idx'] = range(len(dataframe)) return dataframe # ================================================================ # ENTRY SIGNALS # ================================================================ # # Design philosophy: MULTIPLE independent entry signals, each with # only 2-3 conditions. This generates thousands of trades over 13 # months on 15m timeframe, giving hyperopt enough data to optimize. # # Long signals (OR logic - any one triggers entry): # 1. DC bounce: price in lower 30% of Donchian + RSI dip + green candle # 2. BB bounce: price below BB lower + RSI oversold # 3. RSI reversal: RSI was oversold and starts rising + green candle # 4. MACD cross: MACD crosses above signal in lower DC half # # Short signals (OR logic): # 1. DC breakdown: price breaks below DC lower # 2. MACD cross down: MACD crosses below signal + price below EMA # # Quality is managed by DCA, trailing stop, and risk management - # not by filtering entries to near-zero. # ================================================================ def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: # -- LONG 1: Donchian lower zone bounce -- # Price in bottom 30% of Donchian channel + RSI not overbought + green candle long_dc = ( (dataframe['dc_position'] < 0.3) & (dataframe['rsi'] < self.rsi_oversold.value) & (dataframe['green_candle']) & (dataframe['volume'] > 0) ) dataframe.loc[long_dc, ['enter_long', 'enter_tag']] = (1, 'long_dc_bounce') # -- LONG 2: Bollinger Band bounce -- # Price touches/breaks BB lower + RSI dipping long_bb = ( (dataframe['bb_pctb'] < 0.05) & (dataframe['rsi'] < self.rsi_oversold.value + 10) & (dataframe['green_candle']) & (dataframe['volume'] > 0) & (dataframe['enter_long'] != 1) # not already triggered ) dataframe.loc[long_bb, ['enter_long', 'enter_tag']] = (1, 'long_bb_bounce') # -- LONG 3: RSI reversal from oversold -- # RSI was below oversold and starts rising = momentum turning long_rsi = ( (dataframe['rsi'] < self.rsi_oversold.value + 5) & (dataframe['rsi_rising']) & (dataframe['rsi'].shift(1) < self.rsi_oversold.value) & (dataframe['green_candle']) & (dataframe['volume'] > 0) & (dataframe['enter_long'] != 1) ) dataframe.loc[long_rsi, ['enter_long', 'enter_tag']] = (1, 'long_rsi_reversal') # -- LONG 4: MACD bullish crossover in lower Donchian zone -- # MACD crosses above signal while price is in lower half of channel long_macd = ( (dataframe['macd_cross_above']) & (dataframe['dc_position'] < 0.5) & (dataframe['volume'] > 0) & (dataframe['enter_long'] != 1) ) dataframe.loc[long_macd, ['enter_long', 'enter_tag']] = (1, 'long_macd_cross') # -- SHORT 1: Donchian breakdown -- # Price breaks below DC lower channel + red candle short_dc = ( (dataframe['dc_breakout_down']) & (dataframe['red_candle']) & (dataframe['volume'] > 0) ) dataframe.loc[short_dc, ['enter_short', 'enter_tag']] = (1, 'short_dc_breakout') # -- SHORT 2: MACD bearish crossover -- # MACD crosses below signal + price below EMA fast short_macd = ( (dataframe['macd_cross_below']) & (dataframe['close'] < dataframe['ema_fast']) & (dataframe['red_candle']) & (dataframe['volume'] > 0) & (dataframe.get('enter_short', 0) != 1) ) dataframe.loc[short_macd, ['enter_short', 'enter_tag']] = (1, 'short_macd_cross') return dataframe # ================================================================ # EXIT SIGNALS # ================================================================ def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: # -- LONG EXIT 1: Price in upper Donchian zone -- long_exit_dc = ( (dataframe['dc_position'] > 0.85) & (dataframe['red_candle']) & (dataframe['volume'] > 0) ) dataframe.loc[long_exit_dc, ['exit_long', 'exit_tag']] = (1, 'long_dc_upper') # -- LONG EXIT 2: RSI overbought -- long_exit_rsi = ( (dataframe['rsi'] > 70) & (dataframe['rsi_falling']) & (dataframe['volume'] > 0) & (dataframe['exit_long'] != 1) ) dataframe.loc[long_exit_rsi, ['exit_long', 'exit_tag']] = (1, 'long_rsi_overbought') # -- LONG EXIT 3: MACD bearish crossover while in profit zone -- long_exit_macd = ( (dataframe['macd_cross_below']) & (dataframe['dc_position'] > 0.5) & (dataframe['volume'] > 0) & (dataframe['exit_long'] != 1) ) dataframe.loc[long_exit_macd, ['exit_long', 'exit_tag']] = (1, 'long_macd_exit') # -- SHORT EXIT: emergency signal if price explodes up -- short_exit = ( (dataframe['close'] > dataframe['dc_upper']) & (dataframe['green_candle']) ) dataframe.loc[short_exit, ['exit_short', 'exit_tag']] = (1, 'short_emergency_signal') return dataframe # ================================================================ # CUSTOM STOPLOSS (ATR-based stepped trailing) # ================================================================ def custom_stoploss( self, pair: str, trade: Trade, current_time: datetime.datetime, current_rate: float, current_profit: float, after_fill: bool, **kwargs ) -> Optional[float]: """ ATR-based stepped trailing stop for Long. Short: no custom stoploss (TP handled in custom_exit, no SL - house money). Steps: - Profit < breakeven_trigger: default stoploss (-0.99) - Profit >= breakeven_trigger: move to breakeven + 0.2% (cover fees) - Profit 1-3%: trail at 1.5x ATR - Profit 3-5%: trail at 1.0x ATR (tighter) - Profit 5%+: trail at 0.7x ATR (very tight) """ if trade.is_short: # Short: let it ride, no SL - stake is profit bucket money return -0.99 # Get ATR for dynamic trailing try: dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe) if dataframe.empty: return -0.99 atr_pct = dataframe.iloc[-1].get('atr_pct', 1.0) / 100.0 # convert to decimal except Exception: atr_pct = 0.01 # fallback 1% atr_mult = float(self.trailing_atr_mult.value) be_trigger = float(self.breakeven_trigger.value) # Step 1: Below breakeven trigger - default wide SL if current_profit < be_trigger: return -0.99 # Step 2: At breakeven trigger - move to breakeven + fee buffer if current_profit < 0.03: # Trail at 1.5x ATR or breakeven, whichever is tighter trail = max(atr_pct * atr_mult, 0.003) # at least 0.3% sl = min(-trail, -0.002) # at least breakeven + 0.2% return sl # Step 3: 3-5% profit - tighter trailing if current_profit < 0.05: trail = max(atr_pct * atr_mult * 0.7, 0.003) return -trail # Step 4: 5%+ profit - very tight trailing trail = max(atr_pct * atr_mult * 0.5, 0.002) return -trail # ================================================================ # CONFIRM TRADE ENTRY (runtime gate for Short + cooldown) # ================================================================ def confirm_trade_entry( self, pair: str, order_type: str, amount: float, rate: float, time_in_force: str, current_time: datetime.datetime, entry_tag: Optional[str], side: str, **kwargs ) -> bool: """ Gate Short entries: only allow if profit bucket > 0. Gate Long entries: check cooldown after loss. """ if side == 'short': profit = self.custom_profit_bucket.get(pair, 0.0) if profit <= 0: logger.info( f"[SHORT_BLOCKED] {pair} - No profit bucket ({profit:.2f}), " f"blocking Short entry" ) return False logger.info( f"[SHORT_ENTRY] {pair} - Profit bucket: {profit:.2f}, " f"allowing Short at {rate:.2f}" ) self.awaiting_short[pair] = False return True # Long: check cooldown after loss if side == 'long': cooldown_until = self.loss_cooldown_until.get(pair) if cooldown_until and current_time < cooldown_until: logger.info( f"[LONG_COOLDOWN] {pair} - Cooling down until {cooldown_until}, " f"blocking Long entry" ) return False self.dca_count[pair] = 0 self.last_buy_price[pair] = rate self.last_dca_candle[pair] = 0 self.partial_profit_taken[pair] = False logger.info(f"[LONG_ENTRY] {pair} - Entry at {rate:.2f}") return True # ================================================================ # CUSTOM STAKE AMOUNT # ================================================================ def custom_stake_amount( self, current_time: datetime.datetime, current_rate: float, proposed_stake: float, min_stake: Optional[float], max_stake: float, leverage: float, entry_tag: Optional[str], side: str, **kwargs ) -> float: """ Long: Divide wallet across DCA positions. Short: Use profit bucket amount (capped). """ try: if side == 'short': pair = kwargs.get('pair', '') profit = self.custom_profit_bucket.get(pair, 0.0) if profit <= 0: return min_stake or 30.0 # Cap short stake to limit risk cap = float(self.short_max_stake.value) short_stake = min(profit * 0.9, cap) short_stake = max(short_stake, min_stake or 30.0) short_stake = min(short_stake, max_stake) logger.info( f"[SHORT_STAKE] {pair} - Stake: {short_stake:.2f} " f"from bucket: {profit:.2f}, cap: {cap:.2f}" ) return short_stake # Long: split wallet for DCA room balance = self.wallets.get_total_stake_amount() num_positions = self.dca_max_count.value + 1 per_position = balance * 0.8 / num_positions final = max(per_position, min_stake or 30.0) final = min(final, max_stake) logger.info( f"[LONG_STAKE] Balance: {balance:.2f}, " f"Positions: {num_positions}, Stake: {final:.2f}" ) return final except Exception as e: logger.error(f"Error in custom_stake_amount: {e}") return proposed_stake # ================================================================ # LEVERAGE # ================================================================ def leverage( self, pair: str, current_time: datetime.datetime, current_rate: float, proposed_leverage: float, max_leverage: float, entry_tag: Optional[str], side: str, **kwargs ) -> float: """ Long: Dynamic 2-10x based on ATR volatility. Short: Fixed 100x. """ if side == 'short': lev = min(100.0, max_leverage) logger.info(f"[LEVERAGE] {pair} Short: {lev}x") return lev # Long: lower leverage when volatility is high try: dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe) if not dataframe.empty: atr_pct = dataframe.iloc[-1].get('atr_pct', 1.0) lev_min = float(self.long_leverage_min.value) lev_max = float(self.long_leverage_max.value) if atr_pct > 3.0: lev = lev_min elif atr_pct < 0.5: lev = lev_max else: # Linear interpolation: high ATR -> low leverage ratio = (3.0 - atr_pct) / 2.5 lev = lev_min + ratio * (lev_max - lev_min) lev = round(min(lev, max_leverage), 1) logger.info(f"[LEVERAGE] {pair} Long: {lev}x (ATR%: {atr_pct:.2f})") return lev except Exception as e: logger.error(f"Error in leverage: {e}") return min(3.0, max_leverage) # ================================================================ # DCA (adjust_trade_position) - Enhanced # ================================================================ def adjust_trade_position( self, trade: Trade, current_time: datetime.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]: """ DCA for Long only: 1. Price must drop by required % from last buy (progressive spacing) 2. Cooling period: min N candles between DCA orders 3. Max drawdown check: stop DCA if too deep """ pair = trade.pair # No DCA for Short if trade.is_short: return None count = self.dca_count.get(pair, 0) max_dca = self.dca_max_count.value if count >= max_dca: return None # Max drawdown check: stop DCA if unrealized loss > 20% if current_profit < -0.20: logger.info( f"[DCA_BLOCKED] {pair} - Drawdown {current_profit:.2%} > 20%, " f"stopping DCA" ) return None # Progressive spacing: each DCA needs a bigger drop base_drop = float(self.dca_base_drop.value) progressive_factor = 1.0 + (count * 0.3) # gentler progression required_drop = base_drop * progressive_factor last_price = self.last_buy_price.get(pair, None) or trade.open_rate required_price = last_price * (1.0 - required_drop) if current_rate >= required_price: return None # Cooling period check try: dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe) if dataframe.empty: return None last_candle = dataframe.iloc[-1] current_candle_idx = int(last_candle.get('candle_idx', 0)) last_dca_idx = self.last_dca_candle.get(pair, 0) cooling = self.dca_cooling_candles.value if (current_candle_idx - last_dca_idx) < cooling: return None except Exception: return None # Calculate DCA stake with multiplier try: available = self.wallets.get_available_stake_amount() multiplier = float(self.dca_multiplier.value) base_stake = trade.stake_amount dca_stake = base_stake * (multiplier ** (count + 1)) / multiplier dca_stake = min(dca_stake, available * 0.5) dca_stake = min(dca_stake, max_stake) dca_stake = max(dca_stake, min_stake or 30.0) except Exception: dca_stake = min_stake or 30.0 # Update state self.last_buy_price[pair] = current_rate self.dca_count[pair] = count + 1 self.last_dca_candle[pair] = current_candle_idx logger.info( f"[DCA] {pair} #{count+1}/{max_dca} - " f"Price: {current_rate:.2f} (last: {last_price:.2f}, " f"drop: {required_drop:.1%}), Stake: {dca_stake:.2f}" ) return dca_stake # ================================================================ # CUSTOM EXIT (Short TP + Long partial profit) # ================================================================ def custom_exit( self, pair: str, trade: Trade, current_time: datetime.datetime, current_rate: float, current_profit: float, **kwargs ) -> Optional[Union[str, bool]]: """ Short: Take profit at target. Long: Partial profit taking at threshold (sell ~50% signal). """ if trade.is_short: tp = float(self.short_take_profit.value) if current_profit >= tp: logger.info( f"[SHORT_TP] {pair} - Profit: {current_profit:.4f}, TP: {tp}" ) return f'short_tp_{current_profit:.4f}' # Long: partial profit taking if not trade.is_short: pp_pct = float(self.partial_profit_pct.value) if not self.partial_profit_taken.get(pair, False) and current_profit >= pp_pct: self.partial_profit_taken[pair] = True logger.info( f"[PARTIAL_TP] {pair} - Profit: {current_profit:.4f}, " f"taking partial at {pp_pct}" ) return f'partial_tp_{current_profit:.4f}' return None # ================================================================ # CONFIRM TRADE EXIT (Profit Bucket Capture + Cooldown) # ================================================================ 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.datetime, **kwargs ) -> bool: """ On Long exit: capture profit into bucket for Short phase. On Short exit: reset state. On Long loss: set cooldown. """ profit = trade.calc_profit(rate) if not trade.is_short: # Long exit if profit > 0: existing = self.custom_profit_bucket.get(pair, 0.0) self.custom_profit_bucket[pair] = existing + profit self.awaiting_short[pair] = True logger.info( f"[PROFIT_CAPTURE] {pair} - Long profit: {profit:.2f}, " f"Total bucket: {self.custom_profit_bucket[pair]:.2f}, " f"Short phase activated" ) else: # Loss: set cooldown (4 hours = 16 candles at 15m) cooldown_hours = 4 self.loss_cooldown_until[pair] = ( current_time + datetime.timedelta(hours=cooldown_hours) ) logger.info( f"[LONG_EXIT_LOSS] {pair} - Loss: {profit:.2f}, " f"Cooldown until {self.loss_cooldown_until[pair]}, " f"No Short activation" ) # Reset DCA state self.dca_count[pair] = 0 self.last_buy_price.pop(pair, None) self.last_dca_candle.pop(pair, None) self.partial_profit_taken.pop(pair, None) else: # Short exit - log and reset bucket_used = self.custom_profit_bucket.get(pair, 0.0) self.custom_profit_bucket[pair] = 0.0 self.awaiting_short[pair] = False logger.info( f"[SHORT_EXIT] {pair} - Profit: {profit:.2f}, " f"Bucket used: {bucket_used:.2f}, State reset" ) return True # ================================================================ # INFORMATIVE PAIRS # ================================================================ def informative_pairs(self) -> List[Tuple[str, str]]: return []