def main_source_training(source_dataset, target_dataset, target_subject, model, params, weights_exp, eval_mode):
hist_f = params["hist"] // freq
save_file = compute_weights_file(model, source_dataset, target_dataset, target_subject, weights_exp)
train, valid, test, scalers = preprocessing_source_multi(source_dataset, target_dataset, target_subject, ph_f,
hist_f, day_len_f)
make_predictions_tl(target_subject, model, params, ph_f, train, valid, test,
eval_mode=eval_mode, fit=True, save_model_file=save_file)
def _compute_features(self, target_subject, split):
"""
Compute the features for a dataset excluding the given target subject, for a given split
:param target_subject: name or number of target subject to exclude during computation
:param split: number of the split
:return: features, domains (class of subject)
"""
train, valid, test, _ = preprocessing_source_multi(self.source_dataset, self.target_dataset, target_subject,
self.ph, self.hist, self.day_len)
file_path = self._compute_file_path(target_subject)
model = self.Model(target_subject, self.ph, self.params, train[split], valid[split], test[split])
features, _ = model.extract_features("valid", file_path)
y = valid[split].loc[:, ["y", "domain"]].values[:, 1]
return features, y
def __init__(self, source_dataset, target_dataset, target_subject,
exp_name, model_name, params):
self.source_dataset = source_dataset
self.target_dataset = target_dataset
self.target_subject = target_subject
self.exp_name = exp_name
self.model_name = model_name
self.Model = locate_model(model_name)
self.params = locate_params(params)
self.ph = misc.constants.ph_f
self.hist = self.params["hist"] // misc.constants.freq
self.day_len = misc.constants.day_len_f
self.train, self.valid, self.test, _ = preprocessing_source_multi(
self.source_dataset, self.target_dataset, self.target_subject,
self.ph, self.hist, self.day_len)
def end_to_end(source_dataset, target_dataset, target_subject, model, params, weights_exp, eval_mode, exp,
plot):
hist_f = params["hist"] // freq
save_file = compute_weights_file(model, source_dataset, target_dataset, target_subject, weights_exp)
train_m, valid_m, test_m, scalers_m = preprocessing_source_multi(source_dataset, target_dataset, target_subject,
ph_f, hist_f, day_len_f)
make_predictions_tl(target_subject, model, params, ph_f, train_m, valid_m, test_m,
eval_mode=eval_mode, fit=True, save_model_file=save_file)
train, valid, test, scalers = preprocessing(target_dataset, target_subject, ph_f, hist_f, day_len_f)
raw_results = make_predictions_tl(target_subject, model, params, ph_f, train, valid, test,
weights_file=save_file, eval_mode=eval_mode, fit=False, save_model_file=None)
evaluation(raw_results, scalers, source_dataset, target_dataset, target_subject, model, params, exp, plot,
"target_global")
raw_results_2 = make_predictions_tl(target_subject, model, params, ph_f, train, valid, test,
weights_file=save_file, eval_mode=eval_mode, fit=True,
save_model_file=None)
return evaluation(raw_results_2, scalers, source_dataset, target_dataset, target_subject, model, params, exp, plot,
"target_finetuning")