Black beautification.

examples/gp_potentials/cu_training.py

@@ -21,7 +21,7 @@ def train():
         fingerprints=Gaussian(
             cutoff=6.5, normalized=normalized, save_preprocessor="cu_training.scaler"
         ),
-        #model=GaussianProcess(batch_size=batch_size),
+        # model=GaussianProcess(batch_size=batch_size),
         model=GaussianProcess(),
         label="cu_training",
     )

ml4chem/data/handler.py

@@ -90,7 +90,9 @@ def prepare_images(self, images, purpose=None):
         logger.info("Images hashed and processed...\n")
 
         if purpose == "training":
-            logger.info("There are {} atoms in your data set.".format(sum(self.atoms_per_image)))
+            logger.info(
+                "There are {} atoms in your data set.".format(sum(self.atoms_per_image))
+            )
 
     def is_valid_structure(self, images):
         """Check if the data has a valid structure

ml4chem/fingerprints/gaussian.py

@@ -110,7 +110,9 @@ def __init__(
 
         if custom is None:
             custom = {key: custom for key in keys}
-        elif custom is not None and len(list(set(keys).intersection(custom.keys()))) == 0:
+        elif (
+            custom is not None and len(list(set(keys).intersection(custom.keys()))) == 0
+        ):
             for value in custom.values():
                 for k, v in value.items():
                     if isinstance(v, list) is False:
@@ -125,7 +127,7 @@ def __init__(
         # Delete useless variables
         delete = ["self", "scheduler", "overwrite", "k", "v", "value", "keys"]
 
-        for param  in delete:
+        for param in delete:
             try:
                 del _params[param]
             except KeyError:
@@ -218,7 +220,7 @@ def calculate_features(self, images=None, purpose="training", data=None, svm=Fal
             )
             self.custom.update({"GP": self.GP})
         else:
-            logger.info('Using parameters from file to create symmetry functions...\n')
+            logger.info("Using parameters from file to create symmetry functions...\n")
 
         self.print_fingerprint_params(self.GP)
 

ml4chem/metrics.py

@@ -37,6 +37,7 @@ def compute_rmse(outputs, targets, atoms_per_image=None):
     rmse = torch.sqrt(torch.mean((outputs - targets).pow(2))).item()
     return rmse
 
+
 def compute_mse(outputs, targets, atoms_per_image=None):
     """Compute MSE
 

ml4chem/models/gaussian_process.py

@@ -68,6 +68,7 @@ class GaussianProcess(KernelRidge):
         This regressor applies the atomic decomposition Ansatz (ADA). For
         more information check the Notes on the KernelRidge class.
     """
+
     NAME = "GaussianProcess"
 
     def __init__(
@@ -124,7 +125,6 @@ def __init__(
         else:
             self.weights = weights
 
-
     def get_potential_energy(self, fingerprints, reference_space):
         """Get potential energy with Kernel Ridge
 

ml4chem/models/merger.py

@@ -91,7 +91,7 @@ def train(
         batch_size=None,
         lr_scheduler=None,
         independent_loss=True,
-        loss_weights=None
+        loss_weights=None,
     ):
 
         self.epochs = epochs
@@ -107,12 +107,16 @@ def train(
         logging.info("Loss functions:")
 
         if loss_weights is None:
-            self.loss_weights = [1. / len(lossfxn) for l in lossfxn]
+            self.loss_weights = [1.0 / len(lossfxn) for l in lossfxn]
         else:
             self.loss_weights = loss_weights
 
         for index, l in enumerate(lossfxn):
-            logging.info("    - Name: {}; Weight: {}.".format(l.__name__, self.loss_weights[index]))
+            logging.info(
+                "    - Name: {}; Weight: {}.".format(
+                    l.__name__, self.loss_weights[index]
+                )
+            )
 
         # If no batch_size provided then the whole training set length is the batch.
         if batch_size is None:
@@ -192,7 +196,7 @@ def train(
         logging.info("-----------------")
         logging.info("Number of batches:")
         for index, c in enumerate(self.chunks):
-            logging.info('    - Model {}, {}.'.format(index, len(c)))
+            logging.info("    - Model {}, {}.".format(index, len(c)))
         logging.info("Batch size: {} elements per batch.\n".format(batch_size))
 
         # Define optimizer
@@ -207,7 +211,9 @@ def train(
         logger.info(" ")
 
         logger.info(
-            "{:6s} {:19s} {:12s} {:8s}".format("Epoch", "Time Stamp", "Loss", "RMSE (ave)")
+            "{:6s} {:19s} {:12s} {:8s}".format(
+                "Epoch", "Time Stamp", "Loss", "RMSE (ave)"
+            )
         )
         logger.info(
             "{:6s} {:19s} {:12s} {:8s}".format(
@@ -261,13 +267,13 @@ def train(
 
             ts = time.time()
             ts = datetime.datetime.fromtimestamp(ts).strftime("%Y-%m-%d " "%H:%M:%S")
-            logger.info(
-                "{:6d} {} {:8e} {:8f}".format(epoch, ts, loss, _rmse)
-            )
+            logger.info("{:6d} {} {:8e} {:8f}".format(epoch, ts, loss, _rmse))
 
             if convergence is None and epoch == self.epochs:
                 converged = True
-            elif convergence is not None and all(i <= convergence["rmse"] for i in rmse):
+            elif convergence is not None and all(
+                i <= convergence["rmse"] for i in rmse
+            ):
                 converged = True
                 new_state_dict = {}
 
@@ -276,13 +282,12 @@ def train(
 
                 for key in old_state_dict:
                     if not (old_state_dict[key] == new_state_dict[key]).all():
-                        print('Diff in {}'.format(key))
+                        print("Diff in {}".format(key))
                     else:
-                        print('No diff in {}'.format(key))
+                        print("No diff in {}".format(key))
 
             # print(rmse)
 
-
     def closure(self, index, model, independent_loss, name=None):
         """Closure
 

ml4chem/potentials.py

@@ -45,10 +45,9 @@ class Potentials(Calculator, object):
         "PytorchIonicPotentials": "ionic",
         "RetentionTimes": "rt",
         "KernelRidge": "kernelridge",
-        "GaussianProcess": "gaussian_process"
+        "GaussianProcess": "gaussian_process",
     }
 
-
     def __init__(
         self,
         fingerprints=None,
@@ -97,7 +96,9 @@ def load(Cls, model=None, params=None, preprocessor=None, **kwargs):
             class_name = model_params["class_name"]
             module_name = Potentials.module_names[model_params["name"]]
 
-            model_class = dynamic_import(class_name, "ml4chem.models", alt_name=module_name)
+            model_class = dynamic_import(
+                class_name, "ml4chem.models", alt_name=module_name
+            )
 
             delete = ["name", "type", "class_name"]
             for param in delete:
@@ -353,7 +354,9 @@ def calculate(self, atoms, properties, system_changes):
                 try:
                     model.load_state_dict(torch.load(self.ml4chem_path), strict=True)
                 except RuntimeError:
-                    logger.warning('Your image does not have some atoms present in the loaded model.\n')
+                    logger.warning(
+                        "Your image does not have some atoms present in the loaded model.\n"
+                    )
                     model.load_state_dict(torch.load(self.ml4chem_path), strict=False)
                 model.eval()
                 energy = model(fingerprints).item()

ml4chem/utils.py

@@ -156,7 +156,9 @@ def logger(filename=None, level=None, format=None, filemode="a"):
     for handler in logging.root.handlers[:]:
         logging.root.removeHandler(handler)
 
-    logger = logging.basicConfig(filename=filename, level=level, format=format, filemode=filemode)
+    logger = logging.basicConfig(
+        filename=filename, level=level, format=format, filemode=filemode
+    )
 
     return logger
 
@@ -232,9 +234,9 @@ def get_header_message():
           ╚═╝     ╚═╝╚══════╝╚═╝ ╚═════╝╚═╝  ╚═╝╚══════╝╚═╝     ╚═╝\n
 
 
-ML4Chem is Machine Learning for Chemistry. This package is written in Python
-3, and intends to offer modern and rich features to perform machine learning
-workflows for chemical physics.
+ML4Chem is Machine Learning for Chemistry and Materials. This package is
+written in Python 3, and intends to offer modern and rich features to perform
+machine learning workflows for chemical physics.
 
 This project is directed by Muammar El Khatib.
 

setup.py

@@ -24,7 +24,7 @@
     long_description_content_type="text/markdown",
     url="https://github.com/muammar/ml4chem",
     packages=setuptools.find_packages(),
-    scripts=['bin/ml4chem'],
+    scripts=["bin/ml4chem"],
     classifiers=[
         "Programming Language :: Python :: 3",
         "Operating System :: OS Independent",