adelorenzo/stable
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

cleanup tensorboard callback

Browse Source
This commit is contained in:
initrv
2022-12-07 14:37:55 +03:00
parent b9f6911a6a
commit 58604c747e
3 changed files with 21 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
freqtrade/freqai/prediction_models/ReinforcementLearner.py
View File

@@ -100,7 +100,6 @@ class ReinforcementLearner(BaseReinforcementLearningModel):
"""
# first, penalize if the action is not valid
if not self._is_valid(action):
self.custom_info["Invalid"] += 1
return -2
pnl = self.get_unrealized_profit()
@@ -109,15 +108,15 @@ class ReinforcementLearner(BaseReinforcementLearningModel):
# reward agent for entering trades
if (action == Actions.Long_enter.value
and self._position == Positions.Neutral):
self.custom_info[f"{Actions.Long_enter.name}"] += 1
self.custom_info[Actions.Long_enter.name] += 1
return 25
if (action == Actions.Short_enter.value
and self._position == Positions.Neutral):
self.custom_info[f"{Actions.Short_enter.name}"] += 1
self.custom_info[Actions.Short_enter.name] += 1
return 25
# discourage agent from not entering trades
if action == Actions.Neutral.value and self._position == Positions.Neutral:
self.custom_info[f"{Actions.Neutral.name}"] += 1
self.custom_info[Actions.Neutral.name] += 1
return -1
max_trade_duration = self.rl_config.get('max_trade_duration_candles', 300)
@@ -131,22 +130,21 @@ class ReinforcementLearner(BaseReinforcementLearningModel):
# discourage sitting in position
if (self._position in (Positions.Short, Positions.Long) and
action == Actions.Neutral.value):
self.custom_info["Hold"] += 1
self.custom_info[Actions.Neutral.name] += 1
return -1 * trade_duration / max_trade_duration
# close long
if action == Actions.Long_exit.value and self._position == Positions.Long:
if pnl > self.profit_aim * self.rr:
factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)
self.custom_info[f"{Actions.Long_exit.name}"] += 1
self.custom_info[Actions.Long_exit.name] += 1
return float(pnl * factor)
# close short
if action == Actions.Short_exit.value and self._position == Positions.Short:
if pnl > self.profit_aim * self.rr:
factor *= self.rl_config['model_reward_parameters'].get('win_reward_factor', 2)
self.custom_info[f"{Actions.Short_exit.name}"] += 1
self.custom_info[Actions.Short_exit.name] += 1
return float(pnl * factor)
self.custom_info["Unknown"] += 1
return 0.