def load(self, path: Path, details_path: Path = None):
name = FilenameHandler.get_filename(self.__class__.__name__, "pt")
file_path = path / name
if file_path.is_file():
self.model = torch.load(str(file_path))
self.model.eval()
else:
raise FileNotFoundError(
f"{self.__class__.__name__} model could not be loaded from {file_path}. "
f"Check if the path to the {name} file is properly set.")
if details_path is None:
params_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "yaml")
else:
params_path = details_path
if params_path.is_file():
with params_path.open("r") as file:
desc = yaml.safe_load(file)
if "label" in desc:
setattr(self, "label", Label(**desc["label"]))
for param in ["feature_names", "classes"]:
if param in desc:
setattr(self, param, desc[param])
def store(self, path: Path, feature_names=None, details_path=None):
content = self._convert_object_to_dict()
PathBuilder.build(path)
file_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "pickle")
with file_path.open("wb") as file:
pickle.dump(content, file)
if details_path is None:
params_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "yaml")
else:
params_path = details_path
with params_path.open("w") as file:
desc = {
self.label.name: {
**content, "feature_names": feature_names,
"classes": list(self.class_mapping.values())
}
}
if self.label is not None:
desc["label"] = vars(self.label)
yaml.dump(desc, file)
def store(self, path, feature_names=None, details_path: Path = None):
PathBuilder.build(path)
name = FilenameHandler.get_filename(self.__class__.__name__, "pt")
torch.save(self.model, str(path / name))
if details_path is None:
params_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "yaml")
else:
params_path = details_path
with params_path.open("w") as file:
desc = {
**(self.get_params()), "feature_names": feature_names,
"classes": self.get_classes()
}
yaml.dump(desc, file)
def load(self, path: Path):
name = FilenameHandler.get_filename(self.__class__.__name__, "pt")
file_path = path / name
if file_path.is_file():
self.model = torch.load(str(file_path))
self.model.eval()
else:
raise FileNotFoundError(f"{self.__class__.__name__} model could not be loaded from {file_path}. "
f"Check if the path to the {name} file is properly set.")
def run(simulation_state: SimulationState = None):
path = simulation_state.result_path / FilenameHandler.get_dataset_name(
SignalImplanter.__name__)
if path.is_file():
dataset = PickleImport.import_dataset({"path": path},
SignalImplanter.DATASET_NAME)
else:
dataset = SignalImplanter._implant_signals_in_dataset(
simulation_state)
return dataset
def _encode_by_model(self, dataset, params: EncoderParams, vectors):
examples = CacheHandler.memo_by_params(self._prepare_caching_params(dataset, params, vectors,
Word2VecEncoder.DESCRIPTION_REPERTOIRES),
lambda: self._encode_examples(dataset, vectors, params))
if params.encode_labels:
labels = CacheHandler.memo_by_params(self._prepare_caching_params(dataset, params, vectors, Word2VecEncoder.DESCRIPTION_LABELS),
lambda: self._encode_labels(dataset, params))
else:
labels = None
scaler_filename = params.result_path / FilenameHandler.get_filename("standard_scaling", "pkl")
scaled_examples = FeatureScaler.standard_scale(scaler_filename, examples)
encoded_dataset = self._build_encoded_dataset(dataset, scaled_examples, labels, params)
return encoded_dataset
def _vectorize_encoded(self, examples: list, params: EncoderParams):
if self.vectorizer_path is None:
self.vectorizer_path = params.result_path / FilenameHandler.get_filename(DictVectorizer.__name__, "pickle")
if params.learn_model:
vectorizer = DictVectorizer(sparse=True, dtype=float)
vectorized_examples = vectorizer.fit_transform(examples)
PathBuilder.build(params.result_path)
with self.vectorizer_path.open('wb') as file:
pickle.dump(vectorizer, file)
else:
with self.vectorizer_path.open('rb') as file:
vectorizer = pickle.load(file)
vectorized_examples = vectorizer.transform(examples)
return vectorized_examples, vectorizer.get_feature_names()
def load(self, path: Path):
keys = list(vars(self).keys())
file_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "pickle")
if file_path.is_file():
with file_path.open("rb") as file:
content = pickle.load(file)
assert all(
key in keys for key in content.keys()), f"ProbabilisticBinaryClassifier: error while loading from {file_path}: " \
f"object attributes from file and from the class do not match.\n" \
f"Attributes from file: {list(content.keys())}\n" \
f"Attributes for object of class ProbabilisticBinaryClassifier: {keys}"
for key in content:
setattr(self, key, content[key])
else:
raise FileNotFoundError(
f"{self.__class__.__name__} model could not be loaded from {file_path}. "
f"Check if the path to the {file_path.name} file is properly set."
)
def test_get_filename(self):
name = FilenameHandler.get_filename("RandomForestClassifier", "json")
self.assertEqual("random_forest_classifier.json", name)
def test_model_name(self):
name = FilenameHandler.get_model_name("Word2VecEncoder")
self.assertEqual("word2_vec_encoder_model.pickle", name)
def test_get_dataset_name(self):
name = FilenameHandler.get_dataset_name("KmerFrequencyEncoder")
self.assertEqual("encoded_dataset.iml_dataset", name)
def check_if_exists(self, path):
file_path = path / FilenameHandler.get_filename(
self.__class__.__name__, "pt")
return file_path.is_file()
def _get_model_filename(self):
return FilenameHandler.get_filename(self.__class__.__name__, "")