Coverage for source/training/training_config.py: 86%
43 statements
« prev ^ index » next coverage.py v7.8.0, created at 2025-05-31 12:26 +0000
1# training/training_config.py
3from rl.policy import Policy, BoltzmannQPolicy
4from tensorflow.keras.optimizers import Optimizer, Adam
6from ..environment.trading_environment import TradingEnvironment
7from ..agent.agent_handler import AgentHandler
8from ..model.model_blue_prints.base_blue_print import BaseBluePrint
9from ..environment.reward_validator_base import RewardValidatorBase
11class TrainingConfig():
12 """
13 Responsible for creating and configuring training environment and agent.
15 This class encapsulates all configuration parameters needed for training
16 and testing a trading agent, including environment setup, agent creation,
17 and reward validation. It provides a centralized way to manage training
18 parameters and instantiate required components.
19 """
21 def __init__(self, nr_of_steps: int, nr_of_episodes: int, model_blue_print: BaseBluePrint,
22 data_path: str, initial_budget: float, max_amount_of_trades: int, window_size: int,
23 validator: RewardValidatorBase, sell_stop_loss: float = 0.8, sell_take_profit: float = 1.2,
24 buy_stop_loss: float = 0.8, buy_take_profit: float = 1.2, penalty_starts: int = 0,
25 penalty_stops: int = 10, static_reward_adjustment: float = 1, policy: Policy = BoltzmannQPolicy(),
26 optimizer: Optimizer = Adam(learning_rate=1e-3), repeat_test: int = 10, test_ratio: float = 0.2) -> None:
27 """
28 Initializes the training configuration with provided parameters.
30 Parameters:
31 nr_of_steps (int): Total number of training steps.
32 nr_of_episodes (int): Number of training episodes.
33 model_blue_print (BaseBluePrint): Blueprint for creating the neural network model.
34 data_path (str): Path to the training data file.
35 initial_budget (float): Starting budget for the agent.
36 max_amount_of_trades (int): Maximum number of trades allowed to be placed in the environment.
37 window_size (int): Size of the observation window for market data.
38 validator (RewardValidatorBase): Strategy for validating and calculating rewards.
39 sell_stop_loss (float): Coefficient defining when to stop loss on sell positions.
40 sell_take_profit (float): Coefficient defining when to take profit on sell positions.
41 buy_stop_loss (float): Coefficient defining when to stop loss on buy positions.
42 buy_take_profit (float): Coefficient defining when to take profit on buy positions.
43 penalty_starts (int): Starting point (in trading periods without activity) that penalty should be applied from.
44 penalty_stops (int): Ending point (in trading periods without activity) that penalty growth should be stopped at.
45 static_reward_adjustment (float): Adjustment factor for rewards, used to penalize unwanted actions.
46 policy (Policy): Policy for action selection during training.
47 optimizer (Optimizer): Optimizer to be used for model compilation and training.
48 repeat_test (int): Number of times to repeat testing for evaluation.
49 test_ratio (float): Ratio of data to be used for testing vs training.
50 """
52 # Training config
53 self.nr_of_steps = nr_of_steps
54 self.nr_of_episodes = nr_of_episodes
55 self.repeat_test = repeat_test
57 # Environment config
58 self.__data_path: str = data_path
59 self.__test_ratio = test_ratio
60 self.__initial_budget: float = initial_budget
61 self.__max_amount_of_trades: int = max_amount_of_trades
62 self.__window_size: int = window_size
63 self.__sell_stop_loss: float = sell_stop_loss
64 self.__sell_take_profit: float = sell_take_profit
65 self.__buy_stop_loss: float = buy_stop_loss
66 self.__buy_take_profit: float = buy_take_profit
67 self.__penalty_starts: int = penalty_starts
68 self.__penalty_stops: int = penalty_stops
69 self.__static_reward_adjustment: float = static_reward_adjustment
70 self.__validator: RewardValidatorBase = validator
71 self.__instantiated_environment: TradingEnvironment = None
73 # Agent config
74 self.__model_blue_print: BaseBluePrint = model_blue_print
75 self.__policy: Policy = policy
76 self.__optimizer: Optimizer = optimizer
77 self.__instantiated_agent: AgentHandler = None
79 def __str__(self) -> str:
80 """
81 Returns a string representation of the configuration.
83 Creates a formatted multi-line string containing all configuration
84 parameters and their values for easy logging.
86 Returns:
87 str: Formatted string representation of the configuration.
88 """
90 return f"Training config:\n" \
91 f"\tnr_of_steps: {self.nr_of_steps}\n" \
92 f"\tnr_of_episodes: {self.nr_of_episodes}\n" \
93 f"\trepeat_test: {self.repeat_test}\n" \
94 f"\ttest_ratio: {self.__test_ratio}\n" \
95 f"\tinitial_budget: {self.__initial_budget}\n" \
96 f"\tmax_amount_of_trades: {self.__max_amount_of_trades}\n" \
97 f"\twindow_size: {self.__window_size}\n" \
98 f"\tsell_stop_loss: {self.__sell_stop_loss}\n" \
99 f"\tsell_take_profit: {self.__sell_take_profit}\n" \
100 f"\tbuy_stop_loss: {self.__buy_stop_loss}\n" \
101 f"\tbuy_take_profit: {self.__buy_take_profit}\n" \
102 f"\tpenalty_starts: {self.__penalty_starts}\n" \
103 f"\tpenalty_stops: {self.__penalty_stops}\n" \
104 f"\tstatic_reward_adjustment: {self.__static_reward_adjustment}\n" \
105 f"\tvalidator: {self.__validator.__class__.__name__}\n" \
106 f"\t\t{vars(self.__validator)}\n" \
107 f"\tmodel_blue_print: {self.__model_blue_print.__class__.__name__}\n" \
108 f"\t\t{vars(self.__model_blue_print)}\n" \
109 f"\tpolicy: {self.__policy.__class__.__name__}\n" \
110 f"\t\t{vars(self.__policy)}\n" \
111 f"\toptimizer: {self.__optimizer.__class__.__name__}\n" \
112 f"\t\t{self.__optimizer._hyper}\n"
114 def instantiate_environment(self) -> TradingEnvironment:
115 """
116 Creates and returns a TradingEnvironment based on the configuration.
118 Instantiates a new trading environment with the parameters specified
119 in this config. Stores the created environment internally for later use
120 when creating the agent.
122 Returns:
123 TradingEnvironment: Configured trading environment ready for training.
124 """
126 self.__instantiated_environment = TradingEnvironment(self.__data_path,
127 self.__initial_budget,
128 self.__max_amount_of_trades,
129 self.__window_size,
130 self.__validator,
131 self.__sell_stop_loss,
132 self.__sell_take_profit,
133 self.__buy_stop_loss,
134 self.__buy_take_profit,
135 self.__test_ratio,
136 self.__penalty_starts,
137 self.__penalty_stops,
138 self.__static_reward_adjustment)
140 return self.__instantiated_environment
142 def instantiate_agent(self) -> AgentHandler:
143 """
144 Creates and returns an AgentHandler based on the configuration.
146 Uses the model blueprint to create a neural network model with the correct
147 input and output dimensions based on the environment's observation and action
148 spaces. Then wraps this model in an AgentHandler with the specified policy
149 and optimizer.
151 Returns:
152 AgentHandler: Configured agent handler ready for training.
154 Raises:
155 RuntimeError: If environment has not been instantiated first.
156 """
158 if self.__instantiated_environment is None:
159 raise RuntimeError("Environment not instantiated yet!")
161 observation_space_shape = self.__instantiated_environment.observation_space.shape
162 nr_of_actions = self.__instantiated_environment.action_space.n
163 spatial_data_shape = self.__instantiated_environment.get_environment_spatial_data_dimension()
164 model = self.__model_blue_print.instantiate_model(observation_space_shape, nr_of_actions, spatial_data_shape)
165 self.__instantiated_agent = AgentHandler(model, self.__policy, nr_of_actions, self.__optimizer)
167 return self.__instantiated_agent