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

Merge remote-tracking branch 'origin/develop' into update-freqai-tf-handling

Browse Source
This commit is contained in:
Emre
2022-12-11 13:12:45 +03:00
parent 227cdb0938 980a5a9b52
commit 272c3302e3
32 changed files with 460 additions and 237 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
freqtrade/freqai/prediction_models/ReinforcementLearner.py
View File

@@ -71,7 +71,7 @@ class ReinforcementLearner(BaseReinforcementLearningModel):
model.learn(
total_timesteps=int(total_timesteps),
callback=self.eval_callback
callback=[self.eval_callback, self.tensorboard_callback]
)
if Path(dk.data_path / "best_model.zip").is_file():
@@ -100,17 +100,24 @@ 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()
factor = 100.
# reward agent for entering trades
if (action in (Actions.Long_enter.value, Actions.Short_enter.value)
if (action == Actions.Long_enter.value
and self._position == Positions.Neutral):
self.custom_info[f"{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
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
return -1
max_trade_duration = self.rl_config.get('max_trade_duration_candles', 300)
@@ -124,18 +131,22 @@ 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
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
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
return float(pnl * factor)
self.custom_info["Unknown"] += 1
return 0.