Merge pull request #20 from tjkessler/dev

Bug fixes, addition of database creation tools

ecnet/__init__.py

@@ -3,4 +3,4 @@
 import ecnet.error_utils
 import ecnet.model
 import ecnet.limit_parameters
-__version__ = '2.0.0'
+__version__ = '2.0.1'

ecnet/data_utils.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/data_utils.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains the "DataFrame" class, and functions for processing/importing/

ecnet/error_utils.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/error_utils.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains functions for error calculations

ecnet/fitness_functions.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/fitness_functions.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains fitness functions used for input dimensionality reduction,

ecnet/limit_parameters.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/limit_parameters.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains the functions necessary for reducing the input dimensionality of a

ecnet/model.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/error_utils.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains functions necessary creating, training, saving, and reusing neural
@@ -19,7 +19,6 @@
 from copy import deepcopy
 
 # 3rd party imports
-import tensorflow as tf
 from tensorflow import add, global_variables_initializer, matmul, nn
 from tensorflow import placeholder, random_normal, reset_default_graph
 from tensorflow import Session, square, train, Variable
@@ -164,7 +163,7 @@ def fit(self, x_l, y_l, learning_rate=0.1, train_epochs=500,
         sess.close()
 
     def fit_validation(self, x_l, y_l, x_v, y_v, learning_rate=0.1,
-                       max_epochs=10000, keep_prob=0.0):
+                       max_epochs=10000, keep_prob=1.0):
         '''Fits the neural network model using periodic (every 250 epochs)
         validation; if validation set performance does not improve, training
         stops

ecnet/server.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 #
 # ecnet/server.py
-# v.2.0.0
+# v.2.0.1
 # Developed in 2018 by Travis Kessler <[email protected]>
 #
 # Contains the "Server" class, which handles ECNet project creation, neural
@@ -227,7 +227,7 @@ def import_data(self, data_filename, sort_type='random',
             self.DataFrame.create_sets(random=False)
         else:
             raise ValueError('Invalid sort_type {}'.format(sort_type))
-        self.__sets = self.DataFrame.package_sets()
+        self._sets = self.DataFrame.package_sets()
         self._logger.log(
             'info',
             'Imported data from {}'.format(data_filename),
@@ -290,7 +290,7 @@ def limit_input_parameters(self, limit_num, output_filename=None,
                 self.DataFrame, limit_num, self.vars, logger=self._logger
             )
         self.DataFrame.set_inputs(params)
-        self.__sets = self.DataFrame.package_sets()
+        self._sets = self.DataFrame.package_sets()
         if output_filename is not None:
             self.DataFrame.save(output_filename)
             self._logger.log(
@@ -339,7 +339,7 @@ def tune_hyperparameters(self, target_score=None, num_iterations=50,
 
         cost_fn_args = {
             'DataFrame': self.DataFrame,
-            'packaged_data': self.__sets,
+            'packaged_data': self._sets,
             'shuffle': shuffle,
             'data_split': data_split,
             'num_processes': self.__num_processes,
@@ -432,7 +432,7 @@ def train_model(self, validate=False, shuffle=None,
             )
             ecnet.model.train_model(
                 validate,
-                self.__sets,
+                self._sets,
                 self.vars,
                 save_path='./tmp/model'
             )
@@ -465,7 +465,7 @@ def train_model(self, validate=False, shuffle=None,
                                 ecnet.model.train_model,
                                 [
                                     validate,
-                                    self.__sets,
+                                    self._sets,
                                     self.vars,
                                     model_path
                                 ]
@@ -480,14 +480,15 @@ def train_model(self, validate=False, shuffle=None,
                             )
                             ecnet.model.train_model(
                                 validate,
-                                self.__sets,
+                                self._sets,
                                 self.vars,
                                 save_path=model_path
                             )
                         if shuffle is not None:
                             self.DataFrame.shuffle(
                                 shuffle, split=data_split
                             )
+                            self._sets = self.DataFrame.package_sets()
 
             if self.__num_processes > 1:
                 train_pool.close()
@@ -752,7 +753,7 @@ def __open_project(self, project_name):
             self.DataFrame = pload(data_file)
         data_file.close()
 
-        self.__sets = self.DataFrame.package_sets()
+        self._sets = self.DataFrame.package_sets()
         self.__using_project = True
         self._logger.log(
             'info',
@@ -775,26 +776,19 @@ def __determine_x_vals(self, dset):
             'Invalid dset argument: {}'.format(dset)
 
         if dset == 'test':
-            return self.__sets.test_x
+            return self._sets.test_x
         elif dset == 'valid':
-            return self.__sets.valid_x
+            return self._sets.valid_x
         elif dset == 'learn':
-            return self.__sets.learn_x
+            return self._sets.learn_x
         elif dset == 'train':
-            x_vals = []
-            for val in self.__sets.learn_x:
-                x_vals.append(val)
-            for val in self.__sets.valid_x:
-                x_vals.append(val)
+            x_vals = [val for val in self._sets.learn_x]
+            x_vals.extend([val for val in self._sets.valid_x])
             return asarray(x_vals)
         elif dset is None:
-            x_vals = []
-            for val in self.__sets.learn_x:
-                x_vals.append(val)
-            for val in self.__sets.valid_x:
-                x_vals.append(val)
-            for val in self.__sets.test_x:
-                x_vals.append(val)
+            x_vals = [val for val in self._sets.learn_x]
+            x_vals.extend([val for val in self._sets.valid_x])
+            x_vals.extend([val for val in self._sets.test_x])
             return asarray(x_vals)
         else:
             raise ValueError('Unknown dset argument {}'.format(dset))
@@ -814,26 +808,19 @@ def __determine_y_vals(self, dset):
             'Invalid dset argument: {}'.format(dset)
 
         if dset == 'test':
-            return self.__sets.test_y
+            return self._sets.test_y
         elif dset == 'valid':
-            return self.__sets.valid_y
+            return self._sets.valid_y
         elif dset == 'learn':
-            return self.__sets.learn_y
+            return self._sets.learn_y
         elif dset == 'train':
-            y_vals = []
-            for val in self.__sets.learn_y:
-                y_vals.append(val)
-            for val in self.__sets.valid_y:
-                y_vals.append(val)
+            y_vals = [val for val in self._sets.learn_y]
+            y_vals.extend([val for val in self._sets.valid_y])
             return asarray(y_vals)
         elif dset is None:
-            y_vals = []
-            for val in self.__sets.learn_y:
-                y_vals.append(val)
-            for val in self.__sets.valid_y:
-                y_vals.append(val)
-            for val in self.__sets.test_y:
-                y_vals.append(val)
+            y_vals = [val for val in self._sets.learn_y]
+            y_vals.extend([val for val in self._sets.valid_y])
+            y_vals.extend([val for val in self._sets.test_y])
             return asarray(y_vals)
         else:
             raise ValueError('Unknown dset argument {}'.format(dset))

setup.py

@@ -2,7 +2,7 @@
 
 setup(
     name='ecnet',
-    version='2.0.0',
+    version='2.0.1',
     description='UMass Lowell Energy and Combustion Research Laboratory Neural Network Software',
     url='http://github.com/tjkessler/ecnet',
     author='Travis Kessler, Hernan Gelaf-Romer, Sanskriti Sharma',

tools/PaDEL-Descriptor/descriptors.xml

@@ -0,0 +1,86 @@
+<Root>
+    <Group name="2D">
+        <Descriptor name="AcidicGroupCount" value="true"/>
+        <Descriptor name="ALOGP" value="true"/>
+        <Descriptor name="AminoAcidCount" value="false"/>
+        <Descriptor name="APol" value="true"/>
+        <Descriptor name="AromaticAtomsCount" value="true"/>
+        <Descriptor name="AromaticBondsCount" value="true"/>
+        <Descriptor name="AtomCount" value="true"/>
+        <Descriptor name="Autocorrelation" value="true"/>
+        <Descriptor name="BaryszMatrix" value="true"/>
+        <Descriptor name="BasicGroupCount" value="true"/>
+        <Descriptor name="BCUT" value="true"/>
+        <Descriptor name="BondCount" value="true"/>
+        <Descriptor name="BPol" value="true"/>
+        <Descriptor name="BurdenModifiedEigenvalues" value="true"/>
+        <Descriptor name="CarbonTypes" value="true"/>
+        <Descriptor name="ChiChain" value="true"/>
+        <Descriptor name="ChiCluster" value="true"/>
+        <Descriptor name="ChiPathCluster" value="true"/>
+        <Descriptor name="ChiPath" value="true"/>
+        <Descriptor name="Constitutional" value="true"/>
+        <Descriptor name="Crippen" value="true"/>
+        <Descriptor name="DetourMatrix" value="true"/>
+        <Descriptor name="EccentricConnectivityIndex" value="true"/>
+        <Descriptor name="EStateAtomType" value="true"/>
+        <Descriptor name="ExtendedTopochemicalAtom" value="true"/>
+        <Descriptor name="FMF" value="true"/>
+        <Descriptor name="FragmentComplexity" value="true"/>
+        <Descriptor name="HBondAcceptorCount" value="true"/>
+        <Descriptor name="HBondDonorCount" value="true"/>
+        <Descriptor name="HybridizationRatio" value="true"/>
+        <Descriptor name="InformationContent" value="true"/>
+        <Descriptor name="IPMolecularLearning" value="false"/>
+        <Descriptor name="KappaShapeIndices" value="true"/>
+        <Descriptor name="KierHallSmarts" value="false"/>
+        <Descriptor name="LargestChain" value="true"/>
+        <Descriptor name="LargestPiSystem" value="true"/>
+        <Descriptor name="LongestAliphaticChain" value="true"/>
+        <Descriptor name="MannholdLogP" value="true"/>
+        <Descriptor name="McGowanVolume" value="true"/>
+        <Descriptor name="MDE" value="true"/>
+        <Descriptor name="MLFER" value="true"/>
+        <Descriptor name="PathCount" value="true"/>
+        <Descriptor name="PetitjeanNumber" value="true"/>
+        <Descriptor name="RingCount" value="true"/>
+        <Descriptor name="RotatableBondsCount" value="true"/>
+        <Descriptor name="RuleOfFive" value="true"/>
+        <Descriptor name="Topological" value="true"/>
+        <Descriptor name="TopologicalCharge" value="true"/>
+        <Descriptor name="TopologicalDistanceMatrix" value="true"/>
+        <Descriptor name="TPSA" value="true"/>
+        <Descriptor name="VABC" value="true"/>
+        <Descriptor name="VAdjMa" value="true"/>
+        <Descriptor name="WalkCount" value="true"/>
+        <Descriptor name="Weight" value="true"/>
+        <Descriptor name="WeightedPath" value="true"/>
+        <Descriptor name="WienerNumbers" value="true"/>
+        <Descriptor name="XLogP" value="true"/>
+        <Descriptor name="ZagrebIndex" value="true"/>
+    </Group>
+    <Group name="3D">
+        <Descriptor name="Autocorrelation3D" value="true"/>
+        <Descriptor name="CPSA" value="true"/>
+        <Descriptor name="GravitationalIndex" value="true"/>
+        <Descriptor name="LengthOverBreadth" value="true"/>
+        <Descriptor name="MomentOfInertia" value="true"/>
+        <Descriptor name="PetitjeanShapeIndex" value="true"/>
+        <Descriptor name="RDF" value="true"/>
+        <Descriptor name="WHIM" value="true"/>
+    </Group>
+    <Group name="Fingerprint">
+        <Descriptor name="Fingerprinter" value="false"/>
+        <Descriptor name="ExtendedFingerprinter" value="false"/>
+        <Descriptor name="EStateFingerprinter" value="false"/>
+        <Descriptor name="GraphOnlyFingerprinter" value="false"/>
+        <Descriptor name="MACCSFingerprinter" value="false"/>
+        <Descriptor name="PubchemFingerprinter" value="true"/>
+        <Descriptor name="SubstructureFingerprinter" value="false"/>
+        <Descriptor name="SubstructureFingerprintCount" value="false"/>
+        <Descriptor name="KlekotaRothFingerprinter" value="false"/>
+        <Descriptor name="KlekotaRothFingerprintCount" value="false"/>
+        <Descriptor name="AtomPairs2DFingerprinter" value="false"/>
+        <Descriptor name="AtomPairs2DFingerprintCount" value="false"/>
+    </Group>
+</Root>