From: Jérôme Benoit <jerome.benoit@piment-noir.org>
Date: Sat, 1 Mar 2025 11:54:16 +0000 (+0100)
Subject: refactor(reforcexy): revert incorrect change
X-Git-Url: https://git.piment-noir.org/?a=commitdiff_plain;h=a03042300c3a41855375ef3eb1958bce8b103c97;p=freqai-strategies.git

refactor(reforcexy): revert incorrect change

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

diff --git a/ReforceXY/user_data/freqaimodels/ReforceXY.py b/ReforceXY/user_data/freqaimodels/ReforceXY.py
index 2deb6c9..6b68b12 100644
--- a/ReforceXY/user_data/freqaimodels/ReforceXY.py
+++ b/ReforceXY/user_data/freqaimodels/ReforceXY.py
@@ -813,14 +813,14 @@ class ReforceXY(BaseReinforcementLearningModel):
                 self._position in (Positions.Short, Positions.Long)
                 and action == Actions.Neutral.value
             ):
+                max_pnl = max(self.get_most_recent_max_pnl(), pnl)
+                if max_pnl > 0:
+                    drawdown_penalty = 0.01 * factor * (max_pnl - pnl)
+                else:
+                    drawdown_penalty = 0.0
                 lambda1 = 0.05
                 lambda2 = 0.1
                 if pnl >= 0:
-                    max_pnl = max(self.get_most_recent_max_pnl(), pnl)
-                    if max_pnl > 0:
-                        drawdown_penalty = 0.01 * factor * (max_pnl - pnl)
-                    else:
-                        drawdown_penalty = 0.0
                     return (
                         factor
                         * pnl
@@ -829,9 +829,13 @@ class ReforceXY(BaseReinforcementLearningModel):
                         - drawdown_penalty
                     )
                 else:
-                    return factor * pnl * (
-                        1 + lambda1 * (trade_duration / max_trade_duration)
-                    ) - 2 * lambda2 * (trade_duration / max_trade_duration)
+                    return (
+                        factor
+                        * pnl
+                        * (1 + lambda1 * (trade_duration / max_trade_duration))
+                        - 2 * lambda2 * (trade_duration / max_trade_duration)
+                        - drawdown_penalty
+                    )
 
             # close long
             if action == Actions.Long_exit.value and self._position == Positions.Long: