Spaces:
Sleeping
Sleeping
File size: 22,033 Bytes
914ea41 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 |
"""Module containing configuration classes."""
from abc import ABC, abstractmethod
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any, Dict, List, Union
from chython import smarts
import yaml
from CGRtools.containers import MoleculeContainer, QueryContainer
@dataclass
class ConfigABC(ABC):
"""Abstract base class for configuration classes."""
@staticmethod
@abstractmethod
def from_dict(config_dict: Dict[str, Any]):
"""Create an instance of the configuration from a dictionary."""
def to_dict(self) -> Dict[str, Any]:
"""Convert the configuration into a dictionary."""
return {
k: str(v) if isinstance(v, Path) else v for k, v in self.__dict__.items()
}
@staticmethod
@abstractmethod
def from_yaml(file_path: str):
"""Deserialize a YAML file into a configuration object."""
def to_yaml(self, file_path: str):
"""Serializes the configuration to a YAML file.
:param file_path: The path to the output YAML file.
"""
with open(file_path, "w", encoding="utf-8") as file:
yaml.dump(self.to_dict(), file)
@abstractmethod
def _validate_params(self, params: Dict[str, Any]):
"""Validate configuration parameters."""
def __post_init__(self):
"""Validates the configuration parameters."""
# call _validate_params method after initialization
params = self.to_dict()
self._validate_params(params)
@dataclass
class RuleExtractionConfig(ConfigABC):
"""Configuration class for extracting reaction rules.
:param multicenter_rules: If True, extracts a single rule
encompassing all centers. If False, extracts separate reaction
rules for each reaction center in a multicenter reaction.
:param as_query_container: If True, the extracted rules are
generated as QueryContainer objects, analogous to SMARTS objects
for pattern matching in chemical structures.
:param reverse_rule: If True, reverses the direction of the reaction
for rule extraction.
:param reactor_validation: If True, validates each generated rule in
a chemical reactor to ensure correct generation of products from
reactants.
:param include_func_groups: If True, includes specific functional
groups in the reaction rule in addition to the reaction center
and its environment.
:param func_groups_list: A list of functional groups to be
considered when include_func_groups is True.
:param include_rings: If True, includes ring structures in the
reaction rules.
:param keep_leaving_groups: If True, retains leaving groups in the
extracted reaction rule.
:param keep_incoming_groups: If True, retains incoming groups in the
extracted reaction rule.
:param keep_reagents: If True, includes reagents in the extracted
reaction rule.
:param environment_atom_count: Defines the size of the environment
around the reaction center to be included in the rule (0 for
only the reaction center, 1 for the first environment, etc.).
:param min_popularity: Minimum number of times a rule must be
applied to be considered for further analysis.
:param keep_metadata: If True, retains metadata associated with the
reaction in the extracted rule.
:param single_reactant_only: If True, includes only reaction rules
with a single reactant molecule.
:param atom_info_retention: Controls the amount of information about
each atom to retain ('none', 'reaction_center', or 'all').
"""
# default low-level parameters
single_reactant_only: bool = True
keep_metadata: bool = False
reactor_validation: bool = True
reverse_rule: bool = True
as_query_container: bool = True
include_func_groups: bool = False
func_groups_list: List[str] = field(default_factory=list)
# adjustable parameters
environment_atom_count: int = 1
min_popularity: int = 3
include_rings: bool = True
multicenter_rules: bool = True
keep_leaving_groups: bool = True
keep_incoming_groups: bool = True
keep_reagents: bool = False
atom_info_retention: Dict[str, Dict[str, bool]] = field(default_factory=dict)
def __post_init__(self):
super().__post_init__()
self._validate_params(self.to_dict())
self._initialize_default_atom_info_retention()
self._parse_functional_groups()
def _initialize_default_atom_info_retention(self):
default_atom_info = {
"reaction_center": {
"neighbors": True,
"hybridization": True,
"implicit_hydrogens": False,
"ring_sizes": False,
},
"environment": {
"neighbors": False,
"hybridization": False,
"implicit_hydrogens": False,
"ring_sizes": False,
},
}
if not self.atom_info_retention:
self.atom_info_retention = default_atom_info
else:
for key in default_atom_info:
self.atom_info_retention[key].update(
self.atom_info_retention.get(key, {})
)
def _parse_functional_groups(self):
func_groups_list = []
for group_smarts in self.func_groups_list:
try:
query = smarts(group_smarts)
func_groups_list.append(query)
except Exception as e:
print(f"Functional group {group_smarts} was not parsed because of {e}")
self.func_groups_list = func_groups_list
@staticmethod
def from_dict(config_dict: Dict[str, Any]) -> "RuleExtractionConfig":
return RuleExtractionConfig(**config_dict)
@staticmethod
def from_yaml(file_path: str) -> "RuleExtractionConfig":
with open(file_path, "r", encoding="utf-8") as file:
config_dict = yaml.safe_load(file)
return RuleExtractionConfig.from_dict(config_dict)
def _validate_params(self, params: Dict[str, Any]) -> None:
if not isinstance(params["multicenter_rules"], bool):
raise ValueError("multicenter_rules must be a boolean.")
if not isinstance(params["as_query_container"], bool):
raise ValueError("as_query_container must be a boolean.")
if not isinstance(params["reverse_rule"], bool):
raise ValueError("reverse_rule must be a boolean.")
if not isinstance(params["reactor_validation"], bool):
raise ValueError("reactor_validation must be a boolean.")
if not isinstance(params["include_func_groups"], bool):
raise ValueError("include_func_groups must be a boolean.")
if params["func_groups_list"] is not None and not all(
isinstance(group, str) for group in params["func_groups_list"]
):
raise ValueError("func_groups_list must be a list of SMARTS.")
if not isinstance(params["include_rings"], bool):
raise ValueError("include_rings must be a boolean.")
if not isinstance(params["keep_leaving_groups"], bool):
raise ValueError("keep_leaving_groups must be a boolean.")
if not isinstance(params["keep_incoming_groups"], bool):
raise ValueError("keep_incoming_groups must be a boolean.")
if not isinstance(params["keep_reagents"], bool):
raise ValueError("keep_reagents must be a boolean.")
if not isinstance(params["environment_atom_count"], int):
raise ValueError("environment_atom_count must be an integer.")
if not isinstance(params["min_popularity"], int):
raise ValueError("min_popularity must be an integer.")
if not isinstance(params["keep_metadata"], bool):
raise ValueError("keep_metadata must be a boolean.")
if not isinstance(params["single_reactant_only"], bool):
raise ValueError("single_reactant_only must be a boolean.")
if params["atom_info_retention"] is not None:
if not isinstance(params["atom_info_retention"], dict):
raise ValueError("atom_info_retention must be a dictionary.")
required_keys = {"reaction_center", "environment"}
if not required_keys.issubset(params["atom_info_retention"]):
missing_keys = required_keys - set(params["atom_info_retention"].keys())
raise ValueError(
f"atom_info_retention missing required keys: {missing_keys}"
)
for key, value in params["atom_info_retention"].items():
if key not in required_keys:
raise ValueError(f"Unexpected key in atom_info_retention: {key}")
expected_subkeys = {
"neighbors",
"hybridization",
"implicit_hydrogens",
"ring_sizes",
}
if not isinstance(value, dict) or not expected_subkeys.issubset(value):
missing_subkeys = expected_subkeys - set(value.keys())
raise ValueError(
f"Invalid structure for {key} in atom_info_retention. Missing subkeys: {missing_subkeys}"
)
for subkey, subvalue in value.items():
if not isinstance(subvalue, bool):
raise ValueError(
f"Value for {subkey} in {key} of atom_info_retention must be boolean."
)
@dataclass
class PolicyNetworkConfig(ConfigABC):
"""Configuration class for the policy network.
:param vector_dim: Dimension of the input vectors.
:param batch_size: Number of samples per batch.
:param dropout: Dropout rate for regularization.
:param learning_rate: Learning rate for the optimizer.
:param num_conv_layers: Number of convolutional layers in the network.
:param num_epoch: Number of training epochs.
:param policy_type: Mode of operation, either 'filtering' or 'ranking'.
"""
policy_type: str = "ranking"
vector_dim: int = 256
batch_size: int = 500
dropout: float = 0.4
learning_rate: float = 0.008
num_conv_layers: int = 5
num_epoch: int = 100
weights_path: str = None
# for filtering policy
priority_rules_fraction: float = 0.5
rule_prob_threshold: float = 0.0
top_rules: int = 50
@staticmethod
def from_dict(config_dict: Dict[str, Any]) -> "PolicyNetworkConfig":
return PolicyNetworkConfig(**config_dict)
@staticmethod
def from_yaml(file_path: str) -> "PolicyNetworkConfig":
with open(file_path, "r", encoding="utf-8") as file:
config_dict = yaml.safe_load(file)
return PolicyNetworkConfig.from_dict(config_dict)
def _validate_params(self, params: Dict[str, Any]):
if params["policy_type"] not in ["filtering", "ranking"]:
raise ValueError("policy_type must be either 'filtering' or 'ranking'.")
if not isinstance(params["vector_dim"], int) or params["vector_dim"] <= 0:
raise ValueError("vector_dim must be a positive integer.")
if not isinstance(params["batch_size"], int) or params["batch_size"] <= 0:
raise ValueError("batch_size must be a positive integer.")
if (
not isinstance(params["num_conv_layers"], int)
or params["num_conv_layers"] <= 0
):
raise ValueError("num_conv_layers must be a positive integer.")
if not isinstance(params["num_epoch"], int) or params["num_epoch"] <= 0:
raise ValueError("num_epoch must be a positive integer.")
if not isinstance(params["dropout"], float) or not (
0.0 <= params["dropout"] <= 1.0
):
raise ValueError("dropout must be a float between 0.0 and 1.0.")
if (
not isinstance(params["learning_rate"], float)
or params["learning_rate"] <= 0.0
):
raise ValueError("learning_rate must be a positive float.")
if (
not isinstance(params["priority_rules_fraction"], float)
or params["priority_rules_fraction"] < 0.0
):
raise ValueError(
"priority_rules_fraction must be a non-negative positive float."
)
if (
not isinstance(params["rule_prob_threshold"], float)
or params["rule_prob_threshold"] < 0.0
):
raise ValueError("rule_prob_threshold must be a non-negative float.")
if not isinstance(params["top_rules"], int) or params["top_rules"] <= 0:
raise ValueError("top_rules must be a positive integer.")
@dataclass
class ValueNetworkConfig(ConfigABC):
"""Configuration class for the value network.
:param vector_dim: Dimension of the input vectors.
:param batch_size: Number of samples per batch.
:param dropout: Dropout rate for regularization.
:param learning_rate: Learning rate for the optimizer.
:param num_conv_layers: Number of convolutional layers in the network.
:param num_epoch: Number of training epochs.
"""
weights_path: str = None
vector_dim: int = 256
batch_size: int = 500
dropout: float = 0.4
learning_rate: float = 0.008
num_conv_layers: int = 5
num_epoch: int = 100
@staticmethod
def from_dict(config_dict: Dict[str, Any]) -> "ValueNetworkConfig":
return ValueNetworkConfig(**config_dict)
@staticmethod
def from_yaml(file_path: str) -> "ValueNetworkConfig":
with open(file_path, "r", encoding="utf-8") as file:
config_dict = yaml.safe_load(file)
return ValueNetworkConfig.from_dict(config_dict)
def to_yaml(self, file_path: str):
with open(file_path, "w", encoding="utf-8") as file:
yaml.dump(self.to_dict(), file)
def _validate_params(self, params: Dict[str, Any]):
if not isinstance(params["vector_dim"], int) or params["vector_dim"] <= 0:
raise ValueError("vector_dim must be a positive integer.")
if not isinstance(params["batch_size"], int) or params["batch_size"] <= 0:
raise ValueError("batch_size must be a positive integer.")
if (
not isinstance(params["num_conv_layers"], int)
or params["num_conv_layers"] <= 0
):
raise ValueError("num_conv_layers must be a positive integer.")
if not isinstance(params["num_epoch"], int) or params["num_epoch"] <= 0:
raise ValueError("num_epoch must be a positive integer.")
if not isinstance(params["dropout"], float) or not (
0.0 <= params["dropout"] <= 1.0
):
raise ValueError("dropout must be a float between 0.0 and 1.0.")
if (
not isinstance(params["learning_rate"], float)
or params["learning_rate"] <= 0.0
):
raise ValueError("learning_rate must be a positive float.")
@dataclass
class TuningConfig(ConfigABC):
"""Configuration class for the network training.
:param batch_size: The number of targets per batch in the planning simulation step.
:param num_simulations: The number of planning simulations.
"""
batch_size: int = 100
num_simulations: int = 1
@staticmethod
def from_dict(config_dict: Dict[str, Any]) -> "TuningConfig":
return TuningConfig(**config_dict)
@staticmethod
def from_yaml(file_path: str) -> "TuningConfig":
with open(file_path, "r", encoding="utf-8") as file:
config_dict = yaml.safe_load(file)
return TuningConfig.from_dict(config_dict)
def _validate_params(self, params: Dict[str, Any]):
if not isinstance(params["batch_size"], int) or params["batch_size"] <= 0:
raise ValueError("batch_size must be a positive integer.")
@dataclass
class TreeConfig(ConfigABC):
"""Configuration class for the tree search algorithm.
:param max_iterations: The number of iterations to run the algorithm
for.
:param max_tree_size: The maximum number of nodes in the tree.
:param max_time: The time limit (in seconds) for the algorithm to
run.
:param max_depth: The maximum depth of the tree.
:param ucb_type: Type of UCB used in the search algorithm. Options
are "puct", "uct", "value", defaults to "uct".
:param c_ucb: The exploration-exploitation balance coefficient used
in Upper Confidence Bound (UCB).
:param backprop_type: Type of backpropagation algorithm. Options are
"muzero", "cumulative", defaults to "muzero".
:param search_strategy: The strategy used for tree search. Options
are "expansion_first", "evaluation_first".
:param exclude_small: Whether to exclude small molecules during the
search.
:param evaluation_agg: Method for aggregating evaluation scores.
Options are "max", "average", defaults to "max".
:param evaluation_type: The method used for evaluating nodes.
Options are "random", "rollout", "gcn".
:param init_node_value: Initial value for a new node.
:param epsilon: A parameter in the epsilon-greedy search strategy
representing the chance of random selection of reaction rules
during the selection stage in Monte Carlo Tree Search,
specifically during Upper Confidence Bound estimation. It
balances between exploration and exploitation.
:param min_mol_size: Defines the minimum size of a molecule that is
have to be synthesized. Molecules with 6 or fewer heavy atoms
are assumed to be building blocks by definition, thus setting
the threshold for considering larger molecules in the search,
defaults to 6.
:param silent: Whether to suppress progress output.
"""
max_iterations: int = 100
max_tree_size: int = 1000000
max_time: float = 600
max_depth: int = 6
ucb_type: str = "uct"
c_ucb: float = 0.1
backprop_type: str = "muzero"
search_strategy: str = "expansion_first"
exclude_small: bool = True
evaluation_agg: str = "max"
evaluation_type: str = "gcn"
init_node_value: float = 0.0
epsilon: float = 0.0
min_mol_size: int = 6
silent: bool = False
@staticmethod
def from_dict(config_dict: Dict[str, Any]) -> "TreeConfig":
return TreeConfig(**config_dict)
@staticmethod
def from_yaml(file_path: str) -> "TreeConfig":
with open(file_path, "r", encoding="utf-8") as file:
config_dict = yaml.safe_load(file)
return TreeConfig.from_dict(config_dict)
def _validate_params(self, params):
if params["ucb_type"] not in ["puct", "uct", "value"]:
raise ValueError(
"Invalid ucb_type. Allowed values are 'puct', 'uct', 'value'."
)
if params["backprop_type"] not in ["muzero", "cumulative"]:
raise ValueError(
"Invalid backprop_type. Allowed values are 'muzero', 'cumulative'."
)
if params["evaluation_type"] not in ["random", "rollout", "gcn"]:
raise ValueError(
"Invalid evaluation_type. Allowed values are 'random', 'rollout', 'gcn'."
)
if params["evaluation_agg"] not in ["max", "average"]:
raise ValueError(
"Invalid evaluation_agg. Allowed values are 'max', 'average'."
)
if not isinstance(params["c_ucb"], float):
raise TypeError("c_ucb must be a float.")
if not isinstance(params["max_depth"], int) or params["max_depth"] < 1:
raise ValueError("max_depth must be a positive integer.")
if not isinstance(params["max_tree_size"], int) or params["max_tree_size"] < 1:
raise ValueError("max_tree_size must be a positive integer.")
if (
not isinstance(params["max_iterations"], int)
or params["max_iterations"] < 1
):
raise ValueError("max_iterations must be a positive integer.")
if not isinstance(params["max_time"], int) or params["max_time"] < 1:
raise ValueError("max_time must be a positive integer.")
if not isinstance(params["exclude_small"], bool):
raise TypeError("exclude_small must be a boolean.")
if not isinstance(params["silent"], bool):
raise TypeError("silent must be a boolean.")
if not isinstance(params["init_node_value"], float):
raise TypeError("init_node_value must be a float if provided.")
if params["search_strategy"] not in ["expansion_first", "evaluation_first"]:
raise ValueError(
f"Invalid search_strategy: {params['search_strategy']}: "
f"Allowed values are 'expansion_first', 'evaluation_first'"
)
if not isinstance(params["epsilon"], float) or 0 >= params["epsilon"] >= 1:
raise ValueError("epsilon epsilon be a positive float between 0 and 1.")
if not isinstance(params["min_mol_size"], int) or params["min_mol_size"] < 0:
raise ValueError("min_mol_size must be a non-negative integer.")
def convert_config_to_dict(config_attr: ConfigABC, config_type) -> Dict | None:
"""Converts a configuration attribute to a dictionary if it's either a dictionary or
an instance of a specified configuration type.
:param config_attr: The configuration attribute to be converted.
:param config_type: The type to check against for conversion.
:return: The configuration attribute as a dictionary, or None if it's not an
instance of the given type or dict.
"""
if isinstance(config_attr, dict):
return config_attr
if isinstance(config_attr, config_type):
return config_attr.to_dict()
return None
|