using spektral layers

Author: FilippoMB

Autoencoder.py

@@ -6,9 +6,9 @@
 from keras.backend import tf
 from spektral.layers import GraphConvSkip
 from spektral.layers import MinCutPool
+from spektral.utils.convolution import normalized_adjacency
 from utils.misc import sp_matrix_to_sp_tensor_value, get_sw_key
 from tqdm import tqdm
-from spektral.utils.convolution import normalized_adjacency
 from pygsp import graphs
 
 

Clustering.py

@@ -1,4 +1,5 @@
 from collections import OrderedDict
+from tqdm import tqdm
 import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
@@ -15,8 +16,7 @@
 from spektral.layers.convolutional import GraphConvSkip
 from spektral.utils import init_logging
 from spektral.utils.convolution import normalized_adjacency
-from tqdm import tqdm
-from layers import DiffPool, MincutPool
+from spektral.layers import MinCutPool, DiffPool
 from utils import citation
 from utils.misc import sp_matrix_to_sp_tensor_value, product_dict
 np.random.seed(0)  # for reproducibility
@@ -97,7 +97,6 @@
         S_in = Input(tensor=tf.placeholder(tf.int32, shape=(None,), name='segment_ids_in'))
 
         if P['apply_GNN'] and P['method'] != 'diff_pool':
-            print('DEBUG--applying GNN')
             A_norm = normalized_adjacency(A)
             X_1 = GraphConvSkip(P['n_channels'],
                                 kernel_initializer='he_normal',
@@ -107,7 +106,7 @@
             X_1 = X_in
 
         if P['method'] == 'mincut_pool':
-            pool1, adj1, seg1, C = MincutPool(k=n_clust,
+            pool1, adj1, seg1, C = MinCutPool(k=n_clust,
                                               h=P['H_'],
                                               activation=P['ACTIV'])([X_1, A_in, S_in])
 

Graph_Classification.py

@@ -12,10 +12,10 @@
 from keras.regularizers import l2
 from sklearn.model_selection import train_test_split
 from spektral.layers import GraphConv, GlobalAvgPool, ARMAConv, GraphConvSkip
+from spektral.layers import MinCutPool, DiffPool, TopKPool, SAGPool
 from spektral.utils import batch_iterator, log, init_logging
 from spektral.utils.convolution import normalized_adjacency
 from spektral.layers.ops import sp_matrix_to_sp_tensor_value
-from layers import MincutPool, DiffPool, TopKPool, SAGPool
 from utils.dataset_loader import get_graph_kernel_dataset
 
 
@@ -77,7 +77,7 @@ def evaluate(A_list, X_list, y_list, ops):
 # Tunables
 tunables = OrderedDict(
     dataset_ID=['PROTEINS'],
-    method=['flat', 'dense', 'diff_pool', 'top_k_pool', 'mincut_pool', 'sag_pool']
+    method=['mincut_pool'] # 'flat', 'dense', 'diff_pool', 'top_k_pool', 'mincut_pool', 'sag_pool'
 )
 log(tunables)
 
@@ -170,7 +170,7 @@ def evaluate(A_list, X_list, y_list, ops):
             elif P['method'] == 'sag_pool':
                 X_1, A_1, I_1 = SAGPool(0.5)([gc1, A_in, I_in])
             elif P['method'] == 'mincut_pool':
-                X_1, A_1, I_1, M_1 = MincutPool(k=int(average_N // 2),
+                X_1, A_1, I_1, M_1 = MinCutPool(k=int(average_N // 2),
                                                 h=P['mincut_H'],
                                                 activation=P['activ'],
                                                 kernel_regularizer=l2(P['pool_l2']))([gc1, A_in, I_in])
@@ -202,7 +202,7 @@ def evaluate(A_list, X_list, y_list, ops):
             elif P['method'] == 'sag_pool':
                 X_2, A_2, I_2 = SAGPool(0.5)([gc2, A_1, I_1])
             elif P['method'] == 'mincut_pool':
-                X_2, A_2, I_2, M_2 = MincutPool(k=int(average_N // 4),
+                X_2, A_2, I_2, M_2 = MinCutPool(k=int(average_N // 4),
                                                 h=P['mincut_H'],
                                                 activation=P['activ'],
                                                 kernel_regularizer=l2(P['pool_l2']))([gc2, A_1, I_1])

Segmentation.py

@@ -7,7 +7,7 @@
 from skimage.filters import sobel
 from skimage.io import imread
 from keras.models import Model
-from layers import DiffPool, MincutPool
+from spektral.layers import MinCutPool
 from keras import Input
 from keras import backend as K
 from tqdm import tqdm
@@ -123,8 +123,7 @@ def merge_boundary(graph, src, dst):
     A_in = Input(tensor=tf.placeholder(tf.float32, shape=(None, None)), name='A_in')
     S_in = Input(tensor=tf.placeholder(tf.int32, shape=(None,), name='segment_ids_in'))
 
-#    pool1, adj1, seg1, C = DiffPool(n_clust, activation='relu')([X_in,  A_in, S_in])
-    pool1, adj1, seg1, C = MincutPool(n_clust, activation=ACTIV, h=H_)([X_in, A_in, S_in])
+    pool1, adj1, seg1, C = MinCutPool(n_clust, activation=ACTIV, h=H_)([X_in, A_in, S_in])
 
     model = Model([X_in, A_in, S_in], [pool1, seg1])
     model.compile('adam', None)