def save_subdata_sgdml_i(self, i):
from sgdml.utils.io import dataset_md5
ind = self.subdata_indices[i]
name = os.path.join(self.subdata_path, f"subdata_{i}.npz")
data = dict(self.dataset)
for i in ['R', 'E', 'F']:
data[i] = data[i][ind]
data['name'] = name
data['md5'] = dataset_md5(data)
np.savez_compressed(name, **data)
def create_temp_dataset(self, cl_ind):
from sgdml.utils.io import dataset_md5
name = os.path.join(self.storage_dir, f"temp_dataset.npz")
data = dict(self.dataset)
for i in ['R', 'E', 'F']:
data[i] = data[i][cl_ind]
data['name'] = name
data['md5'] = dataset_md5(data)
np.savez_compressed(name, **data)
return name
idxs = model['idxs_' + s]
R = dataset['R'][idxs, :, :]
E = dataset['E'][idxs]
F = dataset['F'][idxs, :, :]
base_vars = {
'type': 'd',
'name': dataset['name'].astype(str),
'theory': dataset['theory'].astype(str),
'z': dataset['z'],
'R': R,
'E': E,
'F': F,
}
base_vars['md5'] = io.dataset_md5(base_vars)
subset_file_name = '%s_%s.npz' % (
os.path.splitext(os.path.basename(dataset_path))[0],
s,
)
file_exists = os.path.isfile(subset_file_name)
if file_exists and args.overwrite:
print(ui.info_str('[INFO]') + ' Overwriting existing model file.')
if not file_exists or args.overwrite:
np.savez_compressed(subset_file_name, **base_vars)
ui.progr_toggle(is_done=True, disp_str='Extracted %s dataset saved to \'%s\'' % (s, subset_file_name))
else:
print(
ui.warn_str('[WARN]')
+ ' %s dataset \'%s\' already exists.' % (s.capitalize(), subset_file_name)
dataset['E'] = dataset['E'][1:25000]
R = dataset['R']
n_dataset = R.shape[0]
print(R.shape)
F_mean = np.zeros(R.shape)
model_dir, model_file_names = args.model_dir
n_models = len(model_file_names)
for i, model_file_name in enumerate(model_file_names):
model_path = os.path.join(model_dir, model_file_name)
print(model_path)
model = np.load(model_path)
gdml = GDMLPredict(model)
_, F = gdml.predict(R.reshape(n_dataset, -1))
F_mean += F.reshape(n_dataset, -1, 3)
F_mean /= n_dataset
dataset = dict(dataset)
dataset['F'] = F_mean
dataset['theory'] = '{} {}'.format('mean_field_F ', dataset['theory'])
dataset['md5'] = io.dataset_md5(dataset)
np.savez_compressed('MEAN_' + dataset_path, **dataset)
