FrictradeLearning.py Stoploss auto et gestion steps de mises adjust

FrictradeLearning.json

@@ -9,22 +9,27 @@
     },
     "trailing": {
       "trailing_stop": false,
-      "trailing_stop_positive": 0.15,
+      "trailing_stop_positive": 0.2,
       "trailing_stop_positive_offset": 1,
       "trailing_only_offset_is_reached": true
     },
     "max_open_trades": {
-      "max_open_trades": 80
-    },
-    "buy": {
-      "hours_force": 11
+      "max_open_trades": 20
     },
     "protection": {
-      "allow_decrease_rate": 0.7,
-      "first_adjust_param": 0.01,
-      "max_steps": 35
+      "allow_decrease_rate": 0.4,
+      "first_adjust_param": 0.005,
+      "max_steps": 45
+    },
+    "buy": {
+      "hours_force": 44,
+      "indic_1h_force_buy": "sma5_deriv1_1h"
+    },
+    "sell": {
+      "offset_max": 18,
+      "offset_min": 17
     }
   },
   "ft_stratparam_v": 1,
-  "export_time": "2025-12-09 07:22:51.929255+00:00"
+  "export_time": "2025-12-14 18:44:03.713386+00:00"
 }

FrictradeLearning.py

@@ -5,77 +5,57 @@
 # IMPORTANT: INSTALL TA BEFOUR RUN(pip install ta)
 # freqtrade hyperopt --hyperopt-loss SharpeHyperOptLoss --spaces buy sell roi --strategy Zeus
 # --- Do not remove these libs ---
-from datetime import timedelta, datetime
-from freqtrade.persistence import Trade
-from freqtrade.strategy import (BooleanParameter, CategoricalParameter, DecimalParameter, stoploss_from_open,
-                                IntParameter, IStrategy, merge_informative_pair, informative, stoploss_from_absolute)
-import pandas as pd
-import numpy as np
-import os
-import json
-import csv
-from pandas import DataFrame
-from typing import Optional, Union, Tuple
-import math
+import inspect
 import logging
-from pathlib import Path
+import os
 
-# --------------------------------
+from datetime import datetime
+from datetime import timezone
+from typing import Optional
 
+import freqtrade.vendor.qtpylib.indicators as qtpylib
+# Machine Learning
+import joblib
+
+import matplotlib.pyplot as plt
+import mpmath as mp
+import numpy as np
+import optuna
+import pandas as pd
+import seaborn as sns
+import shap
 # Add your lib to import here test git
 import ta
 import talib.abstract as talib
-import freqtrade.vendor.qtpylib.indicators as qtpylib
-from datetime import timezone, timedelta
-import mpmath as mp
-
-# Machine Learning
-from sklearn.ensemble import RandomForestClassifier,RandomForestRegressor
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import r2_score, mean_absolute_error, mean_squared_error
-from sklearn.metrics import accuracy_score
-import joblib
-import matplotlib.pyplot as plt
+from freqtrade.persistence import Trade
+from freqtrade.strategy import (CategoricalParameter, DecimalParameter, IntParameter, IStrategy, merge_informative_pair)
+from optuna.visualization import plot_optimization_history
+from optuna.visualization import plot_parallel_coordinate
+from optuna.visualization import plot_param_importances
+from optuna.visualization import plot_slice
+from pandas import DataFrame
+from sklearn.calibration import CalibratedClassifierCV
+from sklearn.feature_selection import SelectFromModel
+from sklearn.feature_selection import VarianceThreshold
+from sklearn.inspection import PartialDependenceDisplay
+from sklearn.inspection import permutation_importance
+from sklearn.linear_model import LogisticRegression
+from sklearn.metrics import brier_score_loss, roc_auc_score
 from sklearn.metrics import (
     classification_report,
     confusion_matrix,
     accuracy_score,
-    roc_auc_score,
     roc_curve,
-    precision_score, recall_score, precision_recall_curve,
-    f1_score
+    precision_score, recall_score
 )
-from sklearn.tree import export_text
-import inspect
-from sklearn.feature_selection import mutual_info_classif
-from sklearn.inspection import permutation_importance
-from lightgbm import LGBMClassifier
-from sklearn.calibration import CalibratedClassifierCV
-from sklearn.feature_selection import SelectFromModel
-from tabulate import tabulate
-from sklearn.model_selection import GridSearchCV
-from sklearn.feature_selection import VarianceThreshold
-import seaborn as sns
-from xgboost import XGBClassifier
-import optuna
-from optuna.visualization import plot_optimization_history
-from optuna.visualization import plot_slice
-from optuna.visualization import plot_param_importances
-from optuna.visualization import plot_parallel_coordinate
-import shap
-from sklearn.inspection import PartialDependenceDisplay
-
-from sklearn.model_selection import train_test_split
 from sklearn.metrics import f1_score
+from sklearn.model_selection import train_test_split
+from sklearn.pipeline import Pipeline
+from sklearn.preprocessing import StandardScaler
+from sklearn.tree import export_text
 from xgboost import XGBClassifier
 
-from sklearn.model_selection import train_test_split
-from sklearn.linear_model import LogisticRegression
-from sklearn.calibration import CalibratedClassifierCV
-from sklearn.metrics import brier_score_loss, roc_auc_score
-
-from sklearn.preprocessing import StandardScaler
-from sklearn.pipeline import Pipeline
+# --------------------------------
 
 
 logger = logging.getLogger(__name__)
@@ -91,7 +71,7 @@ RESET = "\033[0m"
 
 
 class FrictradeLearning(IStrategy):
-    startup_candle_count = 180
+    startup_candle_count = 200
     train_model = None
     model_indicators = []
     DEFAULT_PARAMS = {
@@ -106,13 +86,21 @@ class FrictradeLearning(IStrategy):
         "stoploss": -0.10,
         "minimal_roi": {"0": 0.10}
     }
+    indicators = {'sma24_deriv1', 'sma60_deriv1', 'sma5_deriv1_1h', 'sma12_deriv1_1h', 'sma24_deriv1_1h',
+                  'sma60_deriv1_1h'}
+    indic_1h_force_buy = CategoricalParameter(indicators, default="sma60_deriv1", space='buy')
 
-    allow_decrease_rate = DecimalParameter(0.1, 0.8, decimals=1,  default=0.4, space='protection')
-    first_adjust_param = DecimalParameter(0.001, 0.01, decimals=3,  default=0.003, space='protection')
+    allow_decrease_rate = DecimalParameter(0.1, 0.8, decimals=1, default=0.4, space='protection', optimize=False,
+                                           load=False)
+    first_adjust_param = DecimalParameter(0.001, 0.01, decimals=3, default=0.005, space='protection', optimize=False,
+                                          load=False)
     max_steps = IntParameter(10, 50, default=40, space='protection', optimize=True, load=True)
 
     hours_force = IntParameter(1, 48, default=24, space='buy', optimize=True, load=True)
 
+    offset_min = IntParameter(1, 48, default=24, space='sell', optimize=True, load=True)
+    offset_max = IntParameter(1, 48, default=24, space='sell', optimize=True, load=True)
+
     # ROI table:
     minimal_roi = {
         "0": 10
@@ -157,11 +145,13 @@ class FrictradeLearning(IStrategy):
             'stop': False,
             'max_profit': 0,
             'first_amount': 0,
+            'first_price': 0,
             'total_amount': 0,
             'has_gain': 0,
             'force_sell': False,
             'force_buy': False,
             'last_ath': 0,
+            'mises': {},
             'dca_thresholds': {}
         }
         for pair in ["BTC/USDC", "ETH/USDC", "DOGE/USDC", "XRP/USDC", "SOL/USDC",
@@ -178,11 +168,30 @@ class FrictradeLearning(IStrategy):
         {"date": "2021-11-10", "price_usd": 68742.00, "note": "record novembre 2021 (institutional demand)"},
         {"date": "2024-03-05", "price_usd": 69000.00,
          "note": "nouveau pic début 2024 (source presse, valeur indicative)"},
+        {"date": "2024-03-14", "price_usd": 73816.00,
+         "note": "nouveau pic début 2024 (source presse, valeur indicative)"},
+        {"date": "2024-11-12", "price_usd": 90000.00, "note": ""},
+        {"date": "2024-12-17", "price_usd": 108363.00, "note": ""},
         {"date": "2025-07-11", "price_usd": 118755.00, "note": "pic juillet 2025 (valeur rapportée par la presse)"},
+        {"date": "2025-08-13", "price_usd": 123748.00, "note": ""},
         {"date": "2025-10-06", "price_usd": 126198.07,
          "note": "pic oct. 2025 (source agrégée, à vérifier selon l'exchange)"}
     ]
 
+    def dynamic_trailing_offset(self, price, ath, nb_entries, max_dca=5):
+        dd_ath = (ath - price) / ath
+        dd_ath = max(0.0, min(dd_ath, 0.5))
+
+        dca_risk = min(nb_entries / max_dca, 1.0)
+
+        breathing_score = 0.7 * dd_ath + 0.3 * (1 - dca_risk)
+        breathing_score = min(max(breathing_score, 0.0), 1.0)
+
+        OFFSET_MIN = self.offset_min.value
+        OFFSET_MAX = self.offset_min.value + self.offset_max.value
+
+        return OFFSET_MIN + breathing_score * (OFFSET_MAX - OFFSET_MIN)
+
     def confirm_trade_entry(self, pair: str, order_type: str, amount: float, rate: float, time_in_force: str,
                             current_time: datetime, entry_tag: Optional[str], **kwargs) -> bool:
 
@@ -195,9 +204,9 @@ class FrictradeLearning(IStrategy):
         last_candle_2 = dataframe.iloc[-2].squeeze()
         last_candle_3 = dataframe.iloc[-3].squeeze()
 
-        condition = True #(last_candle[f"{indic_5m}_deriv1"] >= indic_deriv1_5m) and (last_candle[f"{indic_5m}_deriv2"] >= indic_deriv2_5m)
+        condition = True  # (last_candle[f"{indic_5m}_deriv1"] >= indic_deriv1_5m) and (last_candle[f"{indic_5m}_deriv2"] >= indic_deriv2_5m)
 
-        allow_to_buy = True #(condition and not self.pairs[pair]['stop']) | (entry_tag == 'force_entry')
+        allow_to_buy = True  # (condition and not self.pairs[pair]['stop']) | (entry_tag == 'force_entry')
 
         if allow_to_buy:
             self.trades = list()
@@ -235,13 +244,8 @@ class FrictradeLearning(IStrategy):
 
         return allow_to_buy
 
-    def calculateStepsDcaThresholds(self, last_candle, pair):
-        # def split_ratio_one_third(n, p):
-        #     a = n / (2 * p)  # première valeur
-        #     d = n / (p * (p - 1))  # incrément
-        #     return [round(a + i * d, 3) for i in range(p)]
-        def progressive_parts(total, n, first):
-            print('In part')
+    def progressive_parts(self, total, n, first):
+        # print('In part')
         # conditions impossibles → on évite le solveur
         if total <= 0 or first <= 0 or n <= 1:
             return [0] * n
@@ -257,21 +261,31 @@ class FrictradeLearning(IStrategy):
         parts = [round(first * (r ** k), 4) for k in range(n)]
         return parts
 
+    def calculateStepsDcaThresholds(self, last_candle, pair):
+        # def split_ratio_one_third(n, p):
+        #     a = n / (2 * p)  # première valeur
+        #     d = n / (p * (p - 1))  # incrément
+        #     return [round(a + i * d, 3) for i in range(p)]
+
         # r, parts = progressive_parts(0.4, 40, 0.004)
         # print("r =", r)
         # print(parts)
 
-        if self.pairs[pair]['last_ath'] == 0 :
-            ath = max(last_candle['mid'], self.get_last_ath_before_candle(last_candle))
+        val = self.pairs[pair]['first_price'] if self.pairs[pair]['first_price'] > 0 else last_candle['mid']
+
+        if self.pairs[pair]['last_ath'] == 0:
+            ath = max(val, self.get_last_ath_before_candle(last_candle['date']))
             self.pairs[pair]['last_ath'] = ath
 
-        steps = self.approx_value(last_candle['mid'], self.pairs[pair]['last_ath'])
-        self.pairs[pair]['dca_thresholds'] = progressive_parts(
-            (last_candle['mid'] - (self.pairs[pair]['last_ath'] * (1 - self.allow_decrease_rate.value))) / last_candle['mid'],
+        ath = self.pairs[pair]['last_ath']
+
+        steps = self.calculateNumberOfSteps(val, ath, max_steps=self.max_steps.value)
+        self.pairs[pair]['dca_thresholds'] = self.progressive_parts(
+            (val - (ath * (1 - self.allow_decrease_rate.value))) / val,
             steps, self.first_adjust_param.value)
-        print(f"val={last_candle['mid']} lim={self.pairs[pair]['last_ath'] * (1 - self.allow_decrease_rate.value)}"
-              f"steps={steps} "
-              f"pct={(round(last_candle['mid'] - (self.pairs[pair]['last_ath'] * (1 - self.allow_decrease_rate.value))) / last_candle['mid'], 4)}")
+        print(f"val={val} lim={self.pairs[pair]['last_ath'] * (1 - self.allow_decrease_rate.value)}"
+              f" steps={steps}"
+              f" pct={round((val - (self.pairs[pair]['last_ath'] * (1 - self.allow_decrease_rate.value))) / val, 4)}")
         print(self.pairs[pair]['dca_thresholds'])
 
     def confirm_trade_exit(self, pair: str, trade: Trade, order_type: str, amount: float, rate: float,
@@ -283,9 +297,10 @@ class FrictradeLearning(IStrategy):
         last_candle = dataframe.iloc[-1].squeeze()
 
         minutes = int(round((current_time - trade.open_date_utc).seconds / 60, 0))
-        profit =trade.calc_profit(rate)
+        profit = trade.calc_profit(rate)
         force = self.pairs[pair]['force_sell']
-        allow_to_sell = minutes > 30 and (last_candle['hapercent'] < 0 ) or force or (exit_reason == 'force_exit') or (exit_reason == 'stop_loss')
+        # and (last_candle['hapercent'] < 0 )
+        allow_to_sell = True  # (last_candle['hapercent'] < 0 ) or force or (exit_reason == 'force_exit') or (exit_reason == 'stop_loss')
 
         if allow_to_sell:
             self.trades = list()
@@ -306,6 +321,7 @@ class FrictradeLearning(IStrategy):
                 dispo=dispo,
                 profit=round(profit, 2)
             )
+            self.pairs[pair]['first_amount'] = 0
             self.pairs[pair]['max_profit'] = 0
             self.pairs[pair]['force_sell'] = False
             self.pairs[pair]['has_gain'] = 0
@@ -316,8 +332,9 @@ class FrictradeLearning(IStrategy):
             self.pairs[pair]['last_buy'] = 0
             self.pairs[pair]['last_date'] = current_time
             self.pairs[pair]['current_trade'] = None
-        # else:
-            # self.printLog(f"{current_time} SELL triggered for {pair} ({exit_reason} profit={profit} minutes={minutes} percent={last_candle['hapercent']}) but condition blocked")
+        else:
+            self.printLog(
+                f"{current_time} SELL triggered for {pair} ({exit_reason} profit={profit} minutes={minutes} percent={last_candle['hapercent']}) but condition blocked")
         return (allow_to_sell) | (exit_reason == 'force_exit') | (exit_reason == 'stop_loss')
 
     # def custom_exit(self, pair: str, trade: Trade, current_time, current_rate, current_profit, **kwargs):
@@ -387,6 +404,7 @@ class FrictradeLearning(IStrategy):
     def getLastLost(self, last_candle, pair):
         last_lost = round((last_candle['close'] - self.pairs[pair]['max_touch']) / self.pairs[pair]['max_touch'], 3)
         return last_lost
+
     def getPctFirstBuy(self, pair, last_candle):
         return round((last_candle['close'] - self.pairs[pair]['first_buy']) / self.pairs[pair]['first_buy'], 3)
 
@@ -400,7 +418,8 @@ class FrictradeLearning(IStrategy):
             lim = 0.005
             pct = 0.001
         pct_to_max = lim + pct * self.pairs[pair]['count_of_buys']
-        expected_profit = lim * self.pairs[pair]['total_amount']  # min(3 * lim, max(lim, pct_to_max))  # 0.004 + 0.002 * self.pairs[pair]['count_of_buys'] #min(0.01, first_max)
+        expected_profit = lim * self.pairs[pair][
+            'total_amount']  # min(3 * lim, max(lim, pct_to_max))  # 0.004 + 0.002 * self.pairs[pair]['count_of_buys'] #min(0.01, first_max)
 
         self.pairs[pair]['expected_profit'] = expected_profit
 
@@ -415,7 +434,8 @@ class FrictradeLearning(IStrategy):
             self.printLog(
                 f"| {'Date':<16} | {'Action':<10} |{'Pair':<5}| {'Trade Type':<18} |{'Rate':>8} | {'Dispo':>6} | {'Profit':>8} "
                 f"| {'Pct':>6} | {'max_touch':>11} | {'last_lost':>12} | {'last_max':>7}| {'last_min':>7}|{'Buys':>5}| {'Stake':>5} |"
-                f"{'rsi':>6}|{'rsi_1h':>6}|{'rsi_1d':>6}|{'mlprob':>6}" #|Distmax|s201d|s5_1d|s5_2d|s51h|s52h|smt1h|smt2h|tdc1d|tdc1h"
+                f"{'rsi':>6}|{'rsi_1h':>6}|{'rsi_1d':>6}|{'cf_1h':>6}|{'cf_1d':>6}"
+                # |Distmax|s201d|s5_1d|s5_2d|s51h|s52h|smt1h|smt2h|tdc1d|tdc1h"
             )
             self.printLineLog()
             df = pd.DataFrame.from_dict(self.pairs, orient='index')
@@ -467,7 +487,7 @@ class FrictradeLearning(IStrategy):
             f"| {last_max or '-':>7} | {last_min or '-':>7} |{total_counts or '-':>5}|{stake or '-':>7}"
             f"{round(last_candle['max_rsi_24'], 1) or '-':>6}|{round(last_candle['rsi_1h'], 1) or '-':>6}|{round(last_candle['rsi_1d'], 1) or '-':>6}|"
             # f"{round(last_candle['rtp_1h'] * 100, 0) or '-' :>6}|{round(last_candle['rtp_1d'] * 100, 0) or '-' :>6}|"
-            f"{round(last_candle['ml_prob'], 1) or '-':>6}|"
+            # f"{round(last_candle['confidence_index_1d'], 3) or '-':>6}|{round(last_candle['confidence_index_1h'], 3) or '-':>6}|"
         )
 
     def printLineLog(self):
@@ -498,7 +518,8 @@ class FrictradeLearning(IStrategy):
         dataframe['hapercent'] = (dataframe['haclose'] - dataframe['haopen']) / dataframe['haclose']
 
         dataframe['mid'] = dataframe['open'] + (dataframe['close'] - dataframe['open']) / 2
-        dataframe['sma5'] = dataframe['mid'].ewm(span=5, adjust=False).mean() #dataframe["mid"].rolling(window=5).mean()
+        dataframe['sma5'] = dataframe['mid'].ewm(span=5,
+                                                 adjust=False).mean()  # dataframe["mid"].rolling(window=5).mean()
         dataframe['sma5_deriv1'] = 1000 * (dataframe['sma5'] - dataframe['sma5'].shift(1)) / dataframe['sma5'].shift(1)
 
         dataframe['sma12'] = dataframe['mid'].ewm(span=12, adjust=False).mean()
@@ -506,10 +527,12 @@ class FrictradeLearning(IStrategy):
             'sma12'].shift(1)
 
         dataframe['sma24'] = dataframe['mid'].ewm(span=24, adjust=False).mean()
-        dataframe['sma24_deriv1'] = 1000 * (dataframe['sma24'] - dataframe['sma24'].shift(1)) / dataframe['sma24'].shift(1)
+        dataframe['sma24_deriv1'] = 1000 * (dataframe['sma24'] - dataframe['sma24'].shift(1)) / dataframe[
+            'sma24'].shift(1)
 
         dataframe['sma60'] = dataframe['mid'].ewm(span=60, adjust=False).mean()
-        dataframe['sma60_deriv1'] = 1000 * (dataframe['sma60'] - dataframe['sma60'].shift(1)) / dataframe['sma60'].shift(1)
+        dataframe['sma60_deriv1'] = 1000 * (dataframe['sma60'] - dataframe['sma60'].shift(1)) / dataframe[
+            'sma60'].shift(1)
 
         # dataframe[f"sma5_inv"] = (dataframe[f"sma5"].shift(2) >= dataframe[f"sma5"].shift(1)) \
         #                          & (dataframe[f"sma5"].shift(1) <= dataframe[f"sma5"])
@@ -546,7 +569,7 @@ class FrictradeLearning(IStrategy):
         dataframe['max5'] = talib.MAX(dataframe['mid'], timeperiod=5)
         dataframe['min180'] = talib.MIN(dataframe['mid'], timeperiod=180)
         dataframe['max180'] = talib.MAX(dataframe['mid'], timeperiod=180)
-        dataframe['pct180'] = ((dataframe["mid"] - dataframe['min180'] ) / (dataframe['max180'] - dataframe['min180'] ))
+        dataframe['pct180'] = ((dataframe["mid"] - dataframe['min180']) / (dataframe['max180'] - dataframe['min180']))
         dataframe = self.rsi_trend_probability(dataframe, short=60, long=360)
 
         # ################### INFORMATIVE 1h
@@ -564,16 +587,30 @@ class FrictradeLearning(IStrategy):
         informative['macdhist'] = macdhist
 
         informative['rsi'] = talib.RSI(informative['mid'], timeperiod=14)
+
+        informative['sma5'] = informative['mid'].ewm(span=5, adjust=False).mean()
+        informative['sma5_deriv1'] = 1000 * (informative['sma5'] - informative['sma5'].shift(1)) / informative[
+            'sma5'].shift(1)
+
+        informative['sma12'] = informative['mid'].ewm(span=12, adjust=False).mean()
+        informative['sma12_deriv1'] = 1000 * (informative['sma12'] - informative['sma12'].shift(1)) / informative[
+            'sma12'].shift(1)
+
         informative['sma24'] = informative['mid'].ewm(span=24, adjust=False).mean()
-        informative['sma24_deriv1'] = 1000 * (informative['sma24'] - informative['sma24'].shift(1)) / informative['sma24'].shift(1)
+        informative['sma24_deriv1'] = 1000 * (informative['sma24'] - informative['sma24'].shift(1)) / informative[
+            'sma24'].shift(1)
 
         informative['sma60'] = informative['mid'].ewm(span=60, adjust=False).mean()
-        informative['sma60_deriv1'] = 1000 * (informative['sma60'] - informative['sma60'].shift(1)) / informative['sma60'].shift(1)
+        informative['sma60_deriv1'] = 1000 * (informative['sma60'] - informative['sma60'].shift(1)) / informative[
+            'sma60'].shift(1)
+
         informative['rsi'] = talib.RSI(informative['mid'], timeperiod=14)
         self.calculeDerivees(informative, 'rsi', ema_period=12)
         # informative = self.rsi_trend_probability(informative)
 
-        probas = self.calculModelInformative(informative)
+        # probas = self.calculModelInformative(informative)
+
+        # self.calculateConfiance(informative)
 
         # informative = self.populate1hIndicators(df=informative, metadata=metadata)
         # informative = self.calculateRegression(informative, 'mid', lookback=15)
@@ -585,6 +622,8 @@ class FrictradeLearning(IStrategy):
         informative['rsi'] = talib.RSI(informative['mid'], timeperiod=5)
         # informative = self.rsi_trend_probability(informative)
         # informative = self.calculateRegression(informative, 'mid', lookback=15)
+        # self.calculateConfiance(informative)
+
         dataframe = merge_informative_pair(dataframe, informative, '1m', '1d', ffill=True)
 
         dataframe['last_price'] = dataframe['close']
@@ -599,8 +638,8 @@ class FrictradeLearning(IStrategy):
                     filled_buys = trade.select_filled_orders('buy')
                     count = 0
                     amount = 0
-                    min_price = 111111111111110;
-                    max_price = 0;
+                    min_price = 111111111111110
+                    max_price = 0
                     for buy in filled_buys:
                         if count == 0:
                             min_price = min(min_price, buy.price)
@@ -616,7 +655,7 @@ class FrictradeLearning(IStrategy):
                         self.pairs[pair]['last_buy'] = buy.price
                         count = count + 1
                         amount += buy.price * buy.filled
-                    count_buys = count
+                    self.pairs[pair]['count_of_buys'] = count
                     self.pairs[pair]['total_amount'] = amount
 
         dataframe['absolute_min'] = dataframe['mid'].rolling(1440, min_periods=1).min()
@@ -845,11 +884,55 @@ class FrictradeLearning(IStrategy):
         #     d = n / (p * (p - 1))  # incrément
         #     return [round(a + i * d, 3) for i in range(p)]
         #
-        # for val in range(90000, 130000, 10000):
-        #     steps = self.approx_value(val, 126000)
-        #     print(f"val={val} steps={steps} pct={(val - (126000 * (1 - self.allow_decrease_rate.value))) / val}")
-        #     dca   = split_ratio_one_third((val - (126000 * (1 - self.allow_decrease_rate.value))) / 126000, steps)
-        #     print(dca)
+        allow_decrease_rate = 0.3
+        # for val in range(70000, 140000, 10000):
+        #     ath = 126000
+        #
+        #     steps = self.calculateNumberOfSteps(val, ath, max_steps=40)
+        #     self.printLog(f"allow_decrease_rate={self.allow_decrease_rate.value} val={val} steps={steps} pct={round((val - (ath * (1 - allow_decrease_rate))) / val, 4)}")
+        #     # dca   = split_ratio_one_third((val - (ath * (1 - self.allow_decrease_rate.value))) / ath, steps)
+        #     # self.printLog(dca)
+        #     dca_thresholds = self.progressive_parts(
+        #         (val - (ath * (1 - self.allow_decrease_rate.value))) / val,
+        #         steps, self.first_adjust_param.value)
+        #     print(f"val={val} lim={ath * (1 - self.allow_decrease_rate.value)}"
+        #           f"steps={steps} "
+        #           f"pct={(round(val - (ath * (1 - self.allow_decrease_rate.value))) / val, 4)}")
+        #     print(dca_thresholds)
+        ath = 126000
+
+        last_candle = dataframe.iloc[-1].squeeze()
+        val = last_candle['first_price']
+        # steps = self.calculateNumberOfSteps(val, ath, max_steps=40)
+        # self.printLog(
+        #     f"allow_decrease_rate={self.allow_decrease_rate.value} val={val} steps={steps} pct={round((val - (ath * (1 - allow_decrease_rate))) / val, 4)}")
+        # dca_thresholds = self.progressive_parts((val - (ath * (1 - self.allow_decrease_rate.value))) / val, steps, self.first_adjust_param.value)
+        # print(f"val={val} lim={ath * (1 - self.allow_decrease_rate.value)}"
+        #       f"steps={steps} "
+        #       f"pct={(round(val - (ath * (1 - self.allow_decrease_rate.value))) / val, 4)}")
+        # print(dca_thresholds)
+
+
+        if self.dp:
+            if self.dp.runmode.value in ('live', 'dry_run'):
+                full, mises, steps = self.calculateMises(last_candle, pair)
+                # stake = min(self.wallets.get_available_stake_amount(), self.adjust_stake_amount(pair, last_candle))
+                if val and len(self.pairs[pair]['dca_thresholds']) > 0:
+                    print(self.pairs[pair]['dca_thresholds'])
+                    count = 0
+                    pct = 0
+                    dataframe = dataframe.copy()
+                    for dca in self.pairs[pair]['dca_thresholds']:
+                        stake = mises[count]
+                        pct += dca
+                        offset = self.dynamic_trailing_offset(price=val, ath=ath, nb_entries=count)
+
+                        if (count == self.pairs[pair]['count_of_buys']):
+                            print(f"next_buy={val * (1 - pct)} count={count} pct={round(pct, 4)}")
+                            dataframe[f"next_buy"] = val * (1 - pct)
+                        count += 1
+                        print(
+                            f"stake={stake} count={count} pct={round(pct, 4)} offset={offset} next_buy={val * (1 - pct)}")
 
         return dataframe
 
@@ -910,9 +993,9 @@ class FrictradeLearning(IStrategy):
             (dataframe["ml_prob"].shift(1) < dataframe["ml_prob"])
             & (dataframe['sma24_deriv1'] > 0)
             & (dataframe['sma12_deriv1'] > 0)
-            & (dataframe['open'] < dataframe['max180'] * 0.997),
-            # & (dataframe['min180'].shift(3) == dataframe['min180']),
-            ['enter_long', 'enter_tag']
+            # & (dataframe['open'] < dataframe['max180'] * 0.997)
+            # & (dataframe['min180'].shift(3) == dataframe['min180'])
+            , ['enter_long', 'enter_tag']
         ] = (1, f"future")
         dataframe['test'] = np.where(dataframe['enter_long'] == 1, dataframe['close'] * 1.003, np.nan)
 
@@ -1029,50 +1112,40 @@ class FrictradeLearning(IStrategy):
     #
     #     return adjusted_stake_amount
 
-    def approx_value(self, x, X_max):
-        X_min = X_max * (1 - self.allow_decrease_rate.value)  # 126198 * 0.4 = 75718,8
+    def calculateNumberOfSteps(self, current, ath, max_steps=0):
+        if (max_steps == 0):
+            max_steps = self.max_steps.value
+
+        X_min = ath * (1 - self.allow_decrease_rate.value)  # 126198 * 0.4 = 75718,8
         Y_min = 1
-        Y_max = self.max_steps.value
-        a = (Y_max - Y_min) / (X_max - X_min) # 39 ÷ (126198 − 126198×0,6) = 0,000772595
+        Y_max = max_steps
+        a = (Y_max - Y_min) / (ath - X_min)  # 39 ÷ (126198 − 126198×0,6) = 0,000772595
         b = Y_min - a * X_min  # 1 − (0,000772595 × 75718,8) = −38
-        y = a * x + b                         # 0,000772595 * 115000 - 38
+        y = a * current + b  # 0,000772595 * 115000 - 38
         return max(round(y), 1)  # évite les valeurs négatives
 
     def adjust_stake_amount(self, pair: str, last_candle: DataFrame):
 
         if self.pairs[pair]['first_amount'] > 0:
-            return self.pairs[pair]['first_amount']
+            mises = self.pairs[pair]['mises']
+            count = self.pairs[pair]['count_of_buys'] - self.pairs[pair]['has_gain']
+            return mises[count] if count < len(mises) else self.pairs[pair]['first_amount']
 
-        ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle))
+        full, mises, steps = self.calculateMises(last_candle, pair)
+
+        base_stake = mises[self.pairs[pair]['count_of_buys']] if self.pairs[pair]['count_of_buys'] < len(
+            mises) else full / (steps * 2)
+        return base_stake
+
+    def calculateMises(self, last_candle, pair):
+        ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle['date']))
         self.pairs[pair]['last_ath'] = ath
-
-        ath_dist = 100 * (ath - last_candle["mid"]) / ath
-
-
-        # ath_dist
-        #    0  ==> 1
-        #   20  ==> 1.5
-        #   40  ==> 2
-        # 50 * (1 + (ath_dist / 40))
-
         full = self.wallets.get_total_stake_amount()
-        steps = self.approx_value(last_candle['mid'], ath)
-        base_stake = full / steps
-
-        # base_stake = stake * (1 + (ath_dist / 40))
-
-        # Calcule max/min 180
-        low180 = last_candle["min180"]
-        high180 = last_candle["max180"]
-
-        mult = 1 - ((last_candle["mid"] - low180) / (high180 - low180))
-
-        print(f"low={low180} mid={last_candle['mid']} high={high180} mult={mult} ath={ath} ath_dist={round(ath_dist, 2)}" )
-        # base_size = montant de base que tu veux utiliser (ex: stake_amount ou autre)
-        base_size = base_stake  # exemple fraction du portefeuille; adapte selon ton code
-        # new stake proportionnel à mult
-        new_stake = base_size #* mult
-        return new_stake
+        steps = self.calculateNumberOfSteps(last_candle['mid'], ath, max_steps=self.max_steps.value)
+        mises = self.progressive_parts(full, steps, full / (steps * 2))
+        print(f"ath={ath} full={full} steps={steps} mises={mises} ")
+        self.pairs[pair]['mises'] = mises
+        return full, mises, steps
 
     def adjust_trade_position(self, trade: Trade, current_time: datetime,
                               current_rate: float, current_profit: float, min_stake: float,
@@ -1083,6 +1156,11 @@ class FrictradeLearning(IStrategy):
             return None
 
         dataframe, _ = self.dp.get_analyzed_dataframe(trade.pair, self.timeframe)
+
+        if (len(dataframe) < 1):
+            # self.printLog("skip dataframe")
+            return None
+
         last_candle = dataframe.iloc[-1].squeeze()
         # before_last_candle = dataframe.iloc[-2].squeeze()
         # prépare les données
@@ -1099,7 +1177,7 @@ class FrictradeLearning(IStrategy):
         # open_date = trade.open_date.astimezone(timezone.utc)
         # days_since_open = (current_time_utc - open_date).days
         pair = trade.pair
-        profit = trade.calc_profit(current_rate) #round(current_profit * trade.stake_amount, 1)
+        profit = trade.calc_profit(current_rate)  # round(current_profit * trade.stake_amount, 1)
         # last_lost = self.getLastLost(last_candle, pair)
         pct_first = 0
 
@@ -1109,32 +1187,32 @@ class FrictradeLearning(IStrategy):
         # if self.pairs[pair]['first_buy']:
         #     pct_first = self.getPctFirstBuy(pair, last_candle)
 
-        if profit > - self.pairs[pair]['first_amount'] \
-                and self.wallets.get_available_stake_amount() < self.pairs[pair]['first_amount'] \
-                and last_candle['sma24_deriv1_1h'] < 0:
-            stake_amount = trade.stake_amount
-            self.pairs[pair]['previous_profit'] = profit
-            trade_type = "Sell  " + (last_candle['enter_tag'] if last_candle['enter_long'] == 1 else '')
-            self.pairs[trade.pair]['count_of_buys'] += 1
-            self.pairs[pair]['total_amount'] = stake_amount
-            self.log_trade(
-                last_candle=last_candle,
-                date=current_time,
-                action="🟧 Sell +",
-                dispo=dispo,
-                pair=trade.pair,
-                rate=current_rate,
-                trade_type=trade_type,
-                profit=round(profit, 1),
-                buys=trade.nr_of_successful_entries + 1,
-                stake=round(stake_amount, 2)
-            )
-
-            self.pairs[trade.pair]['last_buy'] = current_rate
-            self.pairs[trade.pair]['max_touch'] = last_candle['close']
-            self.pairs[trade.pair]['last_candle'] = last_candle
-
-            return -stake_amount
+        # if profit > - self.pairs[pair]['first_amount'] \
+        #         and self.wallets.get_available_stake_amount() < self.pairs[pair]['first_amount'] \
+        #         and last_candle['sma24_deriv1_1h'] < 0:
+        #     stake_amount = trade.stake_amount
+        #     self.pairs[pair]['previous_profit'] = profit
+        #     trade_type = "Sell  " + (last_candle['enter_tag'] if last_candle['enter_long'] == 1 else '')
+        #     self.pairs[trade.pair]['count_of_buys'] += 1
+        #     self.pairs[pair]['total_amount'] = stake_amount
+        #     self.log_trade(
+        #         last_candle=last_candle,
+        #         date=current_time,
+        #         action="🟥 Stoploss",
+        #         dispo=dispo,
+        #         pair=trade.pair,
+        #         rate=current_rate,
+        #         trade_type=trade_type,
+        #         profit=round(profit, 1),
+        #         buys=trade.nr_of_successful_entries + 1,
+        #         stake=round(stake_amount, 2)
+        #     )
+        #
+        #     self.pairs[trade.pair]['last_buy'] = current_rate
+        #     self.pairs[trade.pair]['max_touch'] = last_candle['close']
+        #     self.pairs[trade.pair]['last_candle'] = last_candle
+        #
+        #     return -stake_amount
 
         if (self.wallets.get_available_stake_amount() < 10):  # or trade.stake_amount >= max_stake:
             return 0
@@ -1145,7 +1223,7 @@ class FrictradeLearning(IStrategy):
             return None
 
         # Dernier prix d'achat réel (pas le prix moyen)
-        last_fill_price = self.pairs[trade.pair]['last_buy'] #trade.open_rate  # remplacé juste après ↓
+        last_fill_price = self.pairs[trade.pair]['last_buy']  # trade.open_rate  # remplacé juste après ↓
 
         # if len(trade.orders) > 0:
         #     # On cherche le dernier BUY exécuté
@@ -1155,7 +1233,7 @@ class FrictradeLearning(IStrategy):
 
         # baisse relative
         # if minutes % 60 == 0:
-        #     ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle))
+        #     ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle['date']))
         #     self.pairs[pair]['last_ath'] = ath
         # else:
         #     ath = self.pairs[pair]['last_ath']
@@ -1166,9 +1244,10 @@ class FrictradeLearning(IStrategy):
         if len(self.pairs[pair]['dca_thresholds']) == 0:
             self.calculateStepsDcaThresholds(last_candle, pair)
 
-        dca_threshold = self.pairs[pair]['dca_thresholds'][min(count_of_buys - 1, len(self.pairs[pair]['dca_thresholds']) - 1)]
+        dca_threshold = self.pairs[pair]['dca_thresholds'][
+            min(count_of_buys - 1, len(self.pairs[pair]['dca_thresholds']) - 1)]
 
-        # print(f"{count_of_buys} {ath * (1 - self.allow_decrease_rate.value)} {round(last_candle['mid'], 2)} {round((last_candle['mid'] - (ath * self.allow_decrease_rate.value)) / last_candle['mid'], 2)} {steps} {round(dca_threshold, 4)}")
+        # self.printLog(f"{count_of_buys} {ath * (1 - self.allow_decrease_rate.value)} {round(last_candle['mid'], 2)} {round((last_candle['mid'] - (ath * self.allow_decrease_rate.value)) / last_candle['mid'], 2)} {steps} {round(dca_threshold, 4)}")
         decline = (last_fill_price - current_rate) / last_fill_price
         increase = - decline
 
@@ -1178,10 +1257,9 @@ class FrictradeLearning(IStrategy):
         #     stake_amount = last_amount * current_rate * 0.5
         #     return stake_amount
 
-
         ########################### ALGO ATH
         # # --- 1. Calcul ATH local de la paire ---
-        # ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle))
+        # ath = max(self.pairs[pair]['last_max'], self.get_last_ath_before_candle(last_candle['date']))
         #
         # # --- 2. Distance ATH - current ---
         # dd = (current_rate - ath) / ath * 100  # dd <= 0
@@ -1208,7 +1286,7 @@ class FrictradeLearning(IStrategy):
         #     return None
 
         # FIN ########################## ALGO ATH
-        force = hours > self.hours_force.value and last_candle['sma60_deriv1_1h'] > 0
+        force = False #hours > self.hours_force.value and last_candle[self.indic_1h_force_buy.value] > 0
         condition = last_candle['percent'] > 0 and last_candle['sma24_deriv1'] > 0 \
                     and last_candle['close'] < self.pairs[pair]['first_buy']
 
@@ -1219,11 +1297,12 @@ class FrictradeLearning(IStrategy):
                     self.pairs[pair]['previous_profit'] = profit
                     return None
 
-                stake_amount = min(self.wallets.get_available_stake_amount(), self.adjust_stake_amount(pair, last_candle))
+                stake_amount = min(self.wallets.get_available_stake_amount(),
+                                   self.adjust_stake_amount(pair, last_candle))
 
-                if force:
-                    stake_amount = stake_amount / 4
-                # print(f"profit={profit} previous={self.pairs[pair]['previous_profit']} count_of_buys={trade.nr_of_successful_entries}")
+                # if force:
+                #     stake_amount = stake_amount / 2
+                # self.printLog(f"profit={profit} previous={self.pairs[pair]['previous_profit']} count_of_buys={trade.nr_of_successful_entries}")
                 if stake_amount > 0:
                     self.pairs[pair]['previous_profit'] = profit
                     trade_type = "Loss " + (last_candle['enter_tag'] if last_candle['enter_long'] == 1 else '')
@@ -1260,11 +1339,14 @@ class FrictradeLearning(IStrategy):
                 self.printLog(exception)
                 return None
 
-        if current_profit > dca_threshold and (increase >= dca_threshold and self.wallets.get_available_stake_amount() > 0)\
+        increase_dca_threshold = 0.003
+        if current_profit > increase_dca_threshold \
+                and (increase >= increase_dca_threshold and self.wallets.get_available_stake_amount() > 0) \
                 and last_candle['rsi'] < 75:
             try:
                 self.pairs[pair]['previous_profit'] = profit
-                stake_amount = max(20, min(self.wallets.get_available_stake_amount(), self.adjust_stake_amount(pair, last_candle)))
+                stake_amount = max(10, min(self.wallets.get_available_stake_amount(),
+                                           self.adjust_stake_amount(pair, last_candle)))
                 if stake_amount > 0:
                     self.pairs[pair]['has_gain'] += 1
 
@@ -1278,7 +1360,7 @@ class FrictradeLearning(IStrategy):
                         dispo=dispo,
                         pair=trade.pair,
                         rate=current_rate,
-                        trade_type='Gain ' + str(round(dca_threshold, 4)),
+                        trade_type='Gain ' + str(round(increase, 4)),
                         profit=round(profit, 1),
                         buys=trade.nr_of_successful_entries + 1,
                         stake=round(stake_amount, 2)
@@ -1304,7 +1386,7 @@ class FrictradeLearning(IStrategy):
         # before_last_candle_12 = dataframe.iloc[-13].squeeze()
         #
         # expected_profit = self.expectedProfit(pair, last_candle)
-        # # print(f"current_time={current_time} current_profit={current_profit} expected_profit={expected_profit}")
+        # # self.printLog(f"current_time={current_time} current_profit={current_profit} expected_profit={expected_profit}")
         #
         # # ----- 1) Charger les variables de trailing pour ce trade -----
         # max_price = self.pairs[pair]['max_touch']
@@ -1315,7 +1397,7 @@ class FrictradeLearning(IStrategy):
 
         count_of_buys = trade.nr_of_successful_entries
 
-        profit = trade.calc_profit(current_rate) #round(current_profit * trade.stake_amount, 1)
+        profit = trade.calc_profit(current_rate)  # round(current_profit * trade.stake_amount, 1)
 
         if current_profit > 0:
             self.pairs[pair]['max_profit'] = max(self.pairs[pair]['max_profit'], profit)
@@ -1325,8 +1407,7 @@ class FrictradeLearning(IStrategy):
         max_profit = self.pairs[pair]['max_profit']
 
         # if current_profit > 0:
-        #     print(f"profit={profit} max_profit={max_profit} current_profit={current_profit}")
-
+        #     self.printLog(f"profit={profit} max_profit={max_profit} current_profit={current_profit}")
 
         # baisse = 0
         # if profit > 0:
@@ -1351,6 +1432,10 @@ class FrictradeLearning(IStrategy):
         current_trailing_only_offset_is_reached = self.trailing_only_offset_is_reached
         current_trailing_stop_positive_offset = self.trailing_stop_positive_offset
 
+        current_trailing_stop_positive_offset = self.dynamic_trailing_offset(price=current_rate,
+                                                                             ath=self.pairs[pair]['last_ath'],
+                                                                             nb_entries=count_of_buys)
+
         # max_ = last_candle['max180']
         # min_ = last_candle['min180']
         # mid = last_candle['mid']
@@ -1393,8 +1478,8 @@ class FrictradeLearning(IStrategy):
         if last_candle['rsi'] > 90:
             return f"rsi_{count_of_buys}_{self.pairs[pair]['has_gain']}"
 
-        if last_candle['sma12'] > last_candle['sma24']: # and last_candle['rsi'] < 85:
-            return None
+        # if last_candle['sma12'] > last_candle['sma24']: # and last_candle['rsi'] < 85:
+        #     return None
 
         # if last_candle['sma24_deriv1'] > 0 : #and minutes < 180 and baisse < 30: # and last_candle['sma5_deriv1'] > -0.15:
         #     if (minutes < 180):
@@ -1402,27 +1487,28 @@ class FrictradeLearning(IStrategy):
         #     if (minutes > 1440 and last_candle['sma60_deriv1'] > 0) :
         #         return None
 
-
         # ----- 4) OFFSET : faut-il attendre de dépasser trailing_stop_positive_offset ? -----
         if current_trailing_only_offset_is_reached and max_profit > current_trailing_stop_positive_offset:
             # Max profit pas atteint ET perte < 2 * current_trailing_stop_positive
-            if profit > max_profit * (1 - current_trailing_stop_positive): #2 * current_trailing_stop_positive:
-                print(f"{current_time} trailing non atteint trailing_stop={round(trailing_stop,4)} profit={round(profit, 4)} max={round(max_profit, 4)} "
-                      f"{min(2, max_profit * (1 - current_trailing_stop_positive))}")
+            if profit > max_profit * (1 - current_trailing_stop_positive):  # 2 * current_trailing_stop_positive:
+                print(
+                    f"{current_time} trailing non atteint trailing_stop={round(trailing_stop, 4)} profit={round(profit, 4)} "
+                    f"max={round(max_profit, 4)} offset={current_trailing_stop_positive_offset}")
                 return None  # ne pas activer le trailing encore
             else:
-                print(f"{current_time} trailing atteint trailing_stop={round(trailing_stop,4)} profit={round(profit, 4)} max={round(max_profit, 4)} "
-                      f"{min(2,max_profit * (1 - current_trailing_stop_positive))}")
+                print(
+                    f"{current_time} trailing atteint trailing_stop={round(trailing_stop, 4)} profit={round(profit, 4)} "
+                    f"max={round(max_profit, 4)} offset={current_trailing_stop_positive_offset}")
         else:
             return None
         # Sinon : trailing actif dès le début
 
         # ----- 6) Condition de vente -----
-        if 0 < profit <= trailing_stop and last_candle['mid'] < last_candle['sma5']:
+        if 0 < profit <= trailing_stop:  # and last_candle['mid'] < last_candle['sma5']: # and profit > current_trailing_stop_positive_offset:
             self.pairs[pair]['force_buy'] = True
             print(
-                f"{current_time} Condition de vente trailing_stop={round(trailing_stop,4)} profit={round(profit, 4)} max={round(max_profit, 4)} "
-                f"{round(min(2, max_profit * (1 - current_trailing_stop_positive)), 4)}")
+                f"{current_time} Condition de vente trailing_stop={round(trailing_stop, 4)} profit={round(profit, 4)} max={round(max_profit, 4)} "
+                f"{round(min(2, max_profit * (1 - current_trailing_stop_positive)), 4)} offset={current_trailing_stop_positive_offset}")
 
             return f"stop_{count_of_buys}_{self.pairs[pair]['has_gain']}"
         return None
@@ -1499,7 +1585,6 @@ class FrictradeLearning(IStrategy):
     #
     #     return df
 
-
     def rsi_trend_probability(self, dataframe, short=6, long=12):
         dataframe = dataframe.copy()
 
@@ -1523,11 +1608,11 @@ class FrictradeLearning(IStrategy):
         return pd.to_datetime(x, utc=True)
 
     # suppose self.btc_ath_history exists (liste de dict)
-    def get_last_ath_before_candle(self, last_candle):
+    def get_last_ath_before_candle(self, date):
 
-        candle_date = self.to_utc_ts(last_candle['date'])  # ou to_utc_ts(last_candle.name)
+        candle_date = self.to_utc_ts(date)  # ou to_utc_ts(last_candle.name)
         best = None
-        for a in self.btc_ath_history: #getattr(self, "btc_ath_history", []):
+        for a in self.btc_ath_history:  # getattr(self, "btc_ath_history", []):
             ath_date = self.to_utc_ts(a["date"])
             if ath_date <= candle_date:
                 if best is None or ath_date > best[0]:
@@ -1539,7 +1624,7 @@ class FrictradeLearning(IStrategy):
         pd.set_option('display.max_rows', None)
         pd.set_option('display.max_columns', None)
         pd.set_option("display.width", 200)
-        path=self.path #f"user_data/plots/{pair}/"
+        path = self.path  # f"user_data/plots/{pair}/"
         os.makedirs(path, exist_ok=True)
 
         # # Étape 1 : sélectionner numériques
@@ -1717,6 +1802,7 @@ class FrictradeLearning(IStrategy):
             best_f1 = max(f1_score(y_valid, (proba > t)) for t in thresholds)
 
             return best_f1
+
         study = optuna.create_study(direction="maximize")
         study.optimize(objective, n_trials=20)
 
@@ -1801,7 +1887,6 @@ class FrictradeLearning(IStrategy):
         fig = plot_parallel_coordinate(study)
         fig.write_html(f"{self.path}/parallel_coordinates.html")
 
-
         # 2️⃣ Sélection des features AVANT calibration
         sfm = SelectFromModel(self.train_model, threshold="median", prefit=True)
         selected_features = X_train.columns[sfm.get_support()]
@@ -1897,7 +1982,6 @@ class FrictradeLearning(IStrategy):
         joblib.dump(self.train_model, f"{self.path}/{pair}_rf_model.pkl")
         print(f"✅ Modèle sauvegardé sous {pair}_rf_model.pkl")
 
-
         # X = dataframe des features (après shift/rolling/indicators)
         # y = target binaire ou décimale
         # model = ton modèle entraîné (RandomForestClassifier ou Regressor)
@@ -2226,15 +2310,19 @@ class FrictradeLearning(IStrategy):
                        and not c.startswith('stop_buying')
                        and not c.startswith('target')
                        and not c.startswith('lvl')
+                       and not c.startswith('sma5_deriv1_1h')
+                       and not c.startswith('sma5_1h')
+                       and not c.startswith('sma12_deriv1_1h')
+                       and not c.startswith('sma12_1h')
+                       and not c.startswith('confidence_index')
                        ]
         # Étape 3 : remplacer inf et NaN par 0
         dataframe[usable_cols] = dataframe[usable_cols].replace([np.inf, -np.inf], 0).fillna(0)
-        print("Colonnes utilisables pour le modèle :")
-        print(usable_cols)
+        # print("Colonnes utilisables pour le modèle :")
+        # print(usable_cols)
         # self.model_indicators = usable_cols
         return usable_cols
 
-
     def select_uncorrelated_features(self, df, target, top_n=20, corr_threshold=0.7):
         """
         Sélectionne les features les plus corrélées avec target,
@@ -2295,11 +2383,12 @@ class FrictradeLearning(IStrategy):
         factor1 = 100 * (ema_period / 5)
         factor2 = 10 * (ema_period / 5)
 
-        dataframe[f"{name}{suffixe}_inv"] = (dataframe[f"{name}{suffixe}"].shift(2) >= dataframe[f"{name}{suffixe}"].shift(1)) \
+        dataframe[f"{name}{suffixe}_inv"] = (dataframe[f"{name}{suffixe}"].shift(2) >= dataframe[
+            f"{name}{suffixe}"].shift(1)) \
                                             & (dataframe[f"{name}{suffixe}"].shift(1) <= dataframe[f"{name}{suffixe}"])
         # --- Distance à la moyenne mobile ---
-        dataframe[f"{name}{suffixe}_dist"] = (dataframe['close'] - dataframe[f"{name}{suffixe}"]) / dataframe[f"{name}{suffixe}"]
-
+        dataframe[f"{name}{suffixe}_dist"] = (dataframe['close'] - dataframe[f"{name}{suffixe}"]) / dataframe[
+            f"{name}{suffixe}"]
 
         # dérivée relative simple
         dataframe[d1_col] = (dataframe[name] - dataframe[name].shift(1)) / dataframe[name].shift(1)
@@ -2380,6 +2469,44 @@ class FrictradeLearning(IStrategy):
 
         return dataframe
 
+    def calculateConfiance(self, informative):
+        df = informative.copy()
+        # ATR normalisé
+        df['atr_norm'] = talib.ATR(df['high'], df['low'], df['close'], length=14) / df['close']
+
+        # SMA200 & pente
+        df['sma200'] = talib.SMA(df['close'], 200)
+        df['sma200_slope'] = df['sma200'].diff()
+
+        # drawdown
+        df['rolling_ath'] = df['close'].cummax()
+        df['drawdown'] = (df['close'] - df['rolling_ath']) / df['rolling_ath']
+
+        # volume spike
+        df['vol_spike'] = df['volume'] / df['volume'].rolling(20).mean()
+
+        # RSI courts/longs
+        df['rsi14'] = talib.RSI(df['close'], 14)
+        df['rsi60'] = talib.RSI(df['close'], 60)
+
+        # Scores normalisés
+        df['vol_score'] = 1 - np.clip(df['atr_norm'] / 0.05, 0, 1)
+        df['trend_score'] = 1 / (1 + np.exp(-df['sma200_slope'] * 150))
+        df['dd_score'] = 1 - np.clip(abs(df['drawdown']) / 0.3, 0, 1)
+        df['volpanic_score'] = 1 - np.clip(df['vol_spike'] / 3, 0, 1)
+        df['rsi_score'] = 1 / (1 + np.exp(-(df['rsi14'] - df['rsi60']) / 10))
+
+        # Indice final
+        informative['confidence_index'] = (
+                0.25 * df['vol_score'] +
+                0.25 * df['trend_score'] +
+                0.20 * df['dd_score'] +
+                0.15 * df['volpanic_score'] +
+                0.15 * df['rsi_score']
+        )
+
+        return informative
+
     def calculModelInformative(self, informative):
         # préparation
         # print(df)
@@ -2415,7 +2542,6 @@ class FrictradeLearning(IStrategy):
         # joindre probabilités au df (dernières lignes correspondantes)
         return probas
 
-
     def prune_features(self, model, dataframe, feature_columns, importance_threshold=0.01):
         """
         Supprime les features dont l'importance est inférieure au seuil.
@@ -2444,6 +2570,6 @@ class FrictradeLearning(IStrategy):
 
         dataframe_pruned = dataframe[kept_features].fillna(0)
 
-        print(f"⚡ Features conservées ({len(kept_features)} / {len(feature_columns)}): {kept_features}")
+        # print(f"⚡ Features conservées ({len(kept_features)} / {len(feature_columns)}): {kept_features}")
 
         return dataframe_pruned, kept_features

plots/BTC/crash/best_threshold.txt

@@ -1 +1 @@
-0.2788135593220339
+0.14152542372881355

tools/sklearn/.ipynb_checkpoints/Sinus-checkpoint.py

@@ -0,0 +1,32 @@
+import numpy as np
+import matplotlib.pyplot as plt
+from lightgbm import LGBMRegressor
+
+# === Données non linéaires ===
+np.random.seed(0)
+X = np.linspace(0, 10, 200).reshape(-1, 1)
+y = np.sin(X).ravel() + np.random.normal(0, 0.1, X.shape[0])  # sinusoïde + bruit
+
+# === Entraînement du modèle ===
+model = LGBMRegressor(
+    n_estimators=300,   # nombre d’arbres
+    learning_rate=0.05, # taux d’apprentissage (plus petit = plus lisse)
+    max_depth=5         # profondeur des arbres (plus grand = plus complexe)
+)
+model.fit(X, y)
+
+# === Prédiction ===
+X_test = np.linspace(0, 10, 500).reshape(-1, 1)
+y_pred = model.predict(X_test)
+
+# === Visualisation ===
+plt.figure(figsize=(10, 5))
+plt.scatter(X, y, color="lightgray", label="Données réelles (sin + bruit)", s=20)
+plt.plot(X_test, np.sin(X_test), color="green", linestyle="--", label="sin(x) réel")
+plt.plot(X_test, y_pred, color="red", label="Prédiction LGBM")
+plt.title("Approximation non linéaire avec LGBMRegressor")
+plt.xlabel("x")
+plt.ylabel("y")
+plt.legend()
+plt.grid(True)
+plt.show()

tools/tensorFlow/Crash.py

@@ -445,6 +445,46 @@ class Crash:
 
         return selected_corr
 
+    def drawPredictions(self, df, indicators, y_proba, threshold=0.5):
+        """
+        Trace simultanément plusieurs indicateurs et la prédiction du modèle.
+
+        Parameters
+        ----------
+        df : pd.DataFrame
+            Dataframe avec les colonnes des indicateurs et la target.
+        indicators : list[str]
+            Liste des colonnes du dataframe à tracer.
+        y_proba : np.array
+            Probabilités prédites par le modèle (valeurs continues entre 0 et 1)
+        threshold : float
+            Seuil pour convertir probabilité en signal binaire
+        output_file : str
+            Fichier sur lequel sauvegarder le graphique
+        """
+        plt.figure(figsize=(18, 6))
+
+        # Tracer les indicateurs
+        for col in indicators:
+            plt.plot(df.index, df[col], label=col, alpha=0.7)
+
+        # Tracer la prédiction du modèle (probabilité)
+        plt.plot(df.index, y_proba, label="Prediction prob.", color="black", linestyle="--")
+
+        # Optionnel : signal binaire (1 si prob > threshold)
+        y_signal = (y_proba > threshold).astype(int)
+        plt.scatter(df.index, y_signal, color='red', marker='o', label='Signal > threshold', s=20)
+
+        plt.title("Indicateurs + prédiction MLP")
+        plt.xlabel("Date")
+        plt.ylabel("Valeur / Probabilité")
+        plt.legend()
+        plt.grid(True)
+        plt.tight_layout()
+        plt.savefig(f"{self.path}/indicators_vs_prediction.png")
+        plt.close()
+
+
     def drawSequentialGraphs(self, model, history, X_train_scaled, X_valid_scaled, y_train,
                              y_valid, thresholds=None, best_threshold=None
                              ):
@@ -556,6 +596,17 @@ class Crash:
         # plt.savefig(f"{self.path}/permutation_importance.png")
         # plt.close()
 
+        # Exemple : on choisit 3 indicateurs du dataframe
+        indicators = ['percent12']
+
+        # y_proba = model.predict(X_valid_scaled).ravel()
+        self.drawPredictions(
+            df=self.dataframe.iloc[-len(y_proba):],  # sélectionner la période correspondant à X_valid
+            indicators=indicators,
+            y_proba=y_proba,
+            threshold=0.5
+        )
+
         # =========================
         # 7️⃣ Sauvegarde seuil
         # =========================
@@ -567,7 +618,7 @@ class Crash:
             "best_f1": max(f1s)
         }
 
-    def optimize_sequential(self, X_train, y_train, X_valid, y_valid, n_trials=20):
+    def optimize_sequential(self, X_train, X_valid, y_train, y_valid , n_trials=20):
         def objective(trial):
             tf.keras.backend.clear_session()
 
@@ -835,7 +886,7 @@ class Crash:
         # df['target'] = df['target'].fillna(0).astype(int)
 
         # label : crash si -n% dans les p heures
-        self.calculateTarget(df)
+        self.initTarget(df)
 
         self.calculateCorrelation(df)
 
@@ -872,7 +923,7 @@ class Crash:
         assert len(X_train) == len(y_train)
         assert len(X_valid) == len(y_valid)
 
-        self.train_model, study = self.optimize_sequential(X_train, y_train, X_valid, y_valid, n_trials=50)
+        self.train_model, study = self.optimize_sequential(X_train, X_valid, y_train, y_valid, n_trials=50)
 
         self.analyseStudy(study)
 
@@ -939,7 +990,7 @@ class Crash:
         # plt.ylabel("Score")
         # plt.show()
 
-        self.analyze_model(pair, self.train_model, X_train, X_valid, y_train, y_valid)
+        self.analyze_model(self.train_model, X_train, X_valid, y_train, y_valid)
 
     def analyseImportances(self, selected_features, X_train, X_valid, y_valid):
         # Feature importance
@@ -1137,7 +1188,7 @@ class Crash:
 
         # FIN SHAP
 
-    def calculateTarget(self, df):
+    def initTarget(self, df):
         future = df['mid'].shift(-12)
         df['future_dd'] = (future - df['mid']) / df['mid']
         df['target'] = (df['future_dd'] > 0.003).astype(int)
@@ -1214,7 +1265,7 @@ class Crash:
 
         print("\n===== ✅ FIN DE L’INSPECTION =====")
 
-    def analyze_model(self, pair, model, X_train, X_valid, y_train, y_valid):
+    def analyze_model(self, model, X_train, X_valid, y_train, y_valid):
         """
         Analyse complète d'un modèle ML supervisé (classification binaire).
         Affiche performances, importance des features, matrices, seuils, etc.
@@ -1398,7 +1449,7 @@ class Crash:
         df = df.set_index('date')
 
         # Optionnel : ne garder qu’une plage temporelle
-        df = df["2025-01-01":"2025-07-14"]
+        df = df["2025-01-01":"2025-02-01"]
         df = df.reset_index('date')
 
         # Supprimer NaN