def test_load(self):
"""Function to test loading a pre-generated model."""
train_features, train_targets, test_features, test_targets = get_data()
self.make_test(train_features, train_targets, test_features,
test_targets)
new_gp = io.read(filename='test-model', ext='pkl')
pred = new_gp.predict(test_fp=test_features,
test_target=test_targets,
get_validation_error=True,
get_training_error=True)
self.assertTrue(
np.allclose(pred['validation_error']['rmse_all'],
self.original['validation_error']['rmse_all']))
gp = io.read(filename='test-model', ext='hdf5')
pred = gp.predict(test_fp=test_features,
test_target=test_targets,
get_validation_error=True,
get_training_error=True)
self.assertTrue(
np.allclose(pred['validation_error']['rmse_all'],
self.original['validation_error']['rmse_all']))
os.remove('{}/test-model.pkl'.format(wkdir))
os.remove('{}/test-model.hdf5'.format(wkdir))
import numpy as np
from sklearn.impute import SimpleImputer
import ase.atoms
from catlearn.fingerprint.adsorbate_prep import autogen_info
from catlearn.fingerprint.setup import FeatureGenerator, default_fingerprinters
from catlearn.regression.gpfunctions import io as gp_io
model_fname = 'apps/catlearn/models/metals_catlearn_gp'
clean_index_name = 'apps/catlearn/train_data/metals_clean_index.npy'
clean_mean = 'apps/catlearn/train_data/metals_clean_feature_mean.npy'
# Load pickled Gaussian process regression model.
gp = gp_io.read(model_fname)
def predict_catkit_demo(images):
"""Return a prediction of adsorption energies for structures generated with
CatKitDemo.
Parameters
----------
images : list
List of atoms objects representing adsorbate-surface structures.
model : str
Path and filename of Catlearn model pickle.
"""
model_ref = {'H': 'H2', 'O': 'H2O, H2', 'C': 'CH4, H2'}
def _get_model():
"""Load the generated model."""
model = gp_io.read('apps/catlearn/models/catapp_catlearn_gp')
return model