perf(qav3): tune SL/TP targets computation

Signed-off-by: Jérôme Benoit <jerome.benoit@piment-noir.org>

diff --git a/quickadapter/user_data/freqaimodels/QuickAdapterRegressorV3.py b/quickadapter/user_data/freqaimodels/QuickAdapterRegressorV3.py
index a2bedc728ea626c332286012e2dd904107ccd9fd..5e83b5e6714fb860d047e5f01d159707a6a05f3b 100644 (file)
--- a/quickadapter/user_data/freqaimodels/QuickAdapterRegressorV3.py
+++ b/quickadapter/user_data/freqaimodels/QuickAdapterRegressorV3.py
@@ -471,10 +471,11 @@ class QuickAdapterRegressorV3(BaseRegressionModel):
             )
             if np_weights.size != normalized_matrix.shape[1]:
                 raise ValueError("label_weights length must match number of objectives")
-            label_knn_metric = self.ft_params.get("label_knn_metric", "euclidean")
             knn_kwargs = {}
+            label_knn_metric = self.ft_params.get("label_knn_metric", "euclidean")
             if label_knn_metric == "minkowski" and isinstance(label_p_order, float):
                 knn_kwargs["p"] = label_p_order
+
             ideal_point = np.ones(normalized_matrix.shape[1])
 
             if metric in {
@@ -587,22 +588,22 @@ class QuickAdapterRegressorV3(BaseRegressionModel):
 
             is_finite_mask = np.isfinite(current_column)
 
-            if is_finite_mask.any():
+            if np.any(is_finite_mask):
                 finite_col = current_column[is_finite_mask]
                 finite_min_val = np.min(finite_col)
                 finite_max_val = np.max(finite_col)
                 finite_range_val = finite_max_val - finite_min_val
 
                 if np.isclose(finite_range_val, 0):
-                    if is_pos_inf_mask.any() and is_neg_inf_mask.any():
+                    if np.any(is_pos_inf_mask) and np.any(is_neg_inf_mask):
                         normalized_matrix[is_finite_mask, i] = 0.5
-                    elif is_pos_inf_mask.any():
+                    elif np.any(is_pos_inf_mask):
                         normalized_matrix[is_finite_mask, i] = (
                             0.0
                             if current_direction == optuna.study.StudyDirection.MAXIMIZE
                             else 1.0
                         )
-                    elif is_neg_inf_mask.any():
+                    elif np.any(is_neg_inf_mask):
                         normalized_matrix[is_finite_mask, i] = (
                             1.0
                             if current_direction == optuna.study.StudyDirection.MAXIMIZE
@@ -1069,7 +1070,7 @@ def zigzag(
     natr_ratio: float = 6.0,
 ) -> tuple[list[int], list[float], list[int]]:
     min_confirmation_window: int = 2
-    max_confirmation_window: int = 5
+    max_confirmation_window: int = 6
     n = len(df)
     if df.empty or n < max(natr_period, 2 * max_confirmation_window + 1):
         return [], [], []
@@ -1446,6 +1447,6 @@ def round_to_nearest_int(value: float, step: int) -> int:
     :return: The rounded value.
     :raises ValueError: If step is zero.
     """
-    if step == 0:
-        raise ValueError("step must be non-zero")
+    if not isinstance(step, int) or step <= 0:
+        raise ValueError("step must be a positive integer")
     return int(round(value / step) * step)

diff --git a/quickadapter/user_data/strategies/QuickAdapterV3.py b/quickadapter/user_data/strategies/QuickAdapterV3.py
index 7cdde0889f4d89ac8d4da5ab1167b07ee12d4497..fb22704867e1e37c20477a03c5babc10d6f7b798 100644 (file)
--- a/quickadapter/user_data/strategies/QuickAdapterV3.py
+++ b/quickadapter/user_data/strategies/QuickAdapterV3.py
@@ -303,7 +303,7 @@ class QuickAdapterV3(IStrategy):
             dataframe["vwap_lowerband"],
             dataframe["vwap_middleband"],
             dataframe["vwap_upperband"],
-        ) = vwapb(dataframe, 20, 1)
+        ) = vwapb(dataframe, 20, 1.0)
         dataframe["%-vwap_width"] = (
             dataframe["vwap_upperband"] - dataframe["vwap_lowerband"]
         ) / dataframe["vwap_middleband"]
@@ -375,10 +375,10 @@ class QuickAdapterV3(IStrategy):
         return self.get_label_natr_ratio(pair) * 0.0125
 
     def get_stoploss_natr_ratio(self, pair: str) -> float:
-        return self.get_label_natr_ratio(pair) * 0.95
+        return self.get_label_natr_ratio(pair) * 0.9
 
     def get_take_profit_natr_ratio(self, pair: str) -> float:
-        return self.get_label_natr_ratio(pair) * 0.75
+        return self.get_label_natr_ratio(pair) * 0.7
 
     def set_freqai_targets(self, dataframe: DataFrame, metadata: dict, **kwargs):
         pair = str(metadata.get("pair"))

diff --git a/quickadapter/user_data/strategies/Utils.py b/quickadapter/user_data/strategies/Utils.py
index 70b163c1f15b6061b08572b705fe82f0f0d226d1..31bd0ccb300cbb6d74922393e9faeb38d4ba61a1 100644 (file)
--- a/quickadapter/user_data/strategies/Utils.py
+++ b/quickadapter/user_data/strategies/Utils.py
@@ -116,11 +116,11 @@ def price_retracement_percent(dataframe: pd.DataFrame, period: int) -> pd.Series
 
 
 # VWAP bands
-def vwapb(dataframe: pd.DataFrame, window=20, num_of_std=1) -> tuple:
+def vwapb(dataframe: pd.DataFrame, window: int = 20, std_factor: float = 1.0) -> tuple:
     vwap = qtpylib.rolling_vwap(dataframe, window=window)
     rolling_std = vwap.rolling(window=window, min_periods=window).std()
-    vwap_low = vwap - (rolling_std * num_of_std)
-    vwap_high = vwap + (rolling_std * num_of_std)
+    vwap_low = vwap - (rolling_std * std_factor)
+    vwap_high = vwap + (rolling_std * std_factor)
     return vwap_low, vwap, vwap_high
 
 
@@ -360,7 +360,7 @@ def zigzag(
     natr_ratio: float = 6.0,
 ) -> tuple[list[int], list[float], list[int]]:
     min_confirmation_window: int = 2
-    max_confirmation_window: int = 5
+    max_confirmation_window: int = 6
     n = len(df)
     if df.empty or n < max(natr_period, 2 * max_confirmation_window + 1):
         return [], [], []