updated property scores

.gitignore

@@ -26,7 +26,7 @@ share/python-wheels/
 .installed.cfg
 *.egg
 MANIFEST
-
+__pycache__
 # PyInstaller
 #  Usually these files are written by a python script from a template
 #  before PyInstaller builds the exe, so as to inject date/other infos into it.

README.md

@@ -5,4 +5,9 @@
 A sample of command to run GNN models. More will be added.
 ```
 python main.py --edgelist datasets/input2f/email.edgelist --label datasets/input2f/email.nodes.labels
+python main.py --edgelist datasets/input2f/usaairports.edgelist --label datasets/input2f/usaairports.nodes.labels --oneindexed 1
+python main.py --edgelist datasets/input2f/yeast.edgelist --label datasets/input2f/yeast.nodes.labels --oneindexed 1 --onelabeled 1
+python main.py --edgelist datasets/input3f/squirrel_edges.txt --label datasets/input3f/squirrel_labels.txt --feature datasets/input3f/squirrel_features.txt
+
+
 ```

main.py

@@ -22,7 +22,7 @@
 import time
 from sklearn.metrics import confusion_matrix
 from sklearn.metrics import accuracy_score
-from utils import readInput3f, readInput2f, loadPyGDataset
+from utils import readInput3f, readInput2f, loadPyGDataset, computeHomophily, mixingCommunityScore
 from markov import (
 markov_process_agg, 
 markov_process_disj, 
@@ -115,4 +115,6 @@ def helper(data, args):
     else:
         data = loadPyGDataset(dataset_name)
     print(data)
+    if args.debug == 1:
+        print("Homophily:", computeHomophily(data).item(), "Community Mixing:", mixingCommunityScore(data))
     helper(data, args)

markov.py

@@ -7,6 +7,8 @@
 from torch_sparse import spspmm
 from torch_geometric.utils import add_remaining_self_loops, to_dense_adj, dense_to_sparse
 from torch_scatter import scatter_add
+from utils import computeHomophily, newEdges
+
 
 def markov_normalization(edge_index, edge_weight, num_nodes, ntype = 'col'):
     if ntype == 'col':
@@ -46,6 +48,9 @@ def markov_process_agg_sparse(data, eps, inflate, nlayers, row_normalization = T
         remaining_edge_idx = torch.nonzero(ew >= eps).flatten()
         ei = ei[:,remaining_edge_idx]
         ew = ew[remaining_edge_idx]
+        if ei.shape[1] < 1:
+            print("No more edges..! stopping after ", i, "layers")
+            break
         # normalization
         if row_normalization:
             ei, ew = markov_normalization(ei, ew, len(data.x), 'row')
@@ -55,7 +60,7 @@ def markov_process_agg_sparse(data, eps, inflate, nlayers, row_normalization = T
         medge_index.append(ei)
         medge_weight.append(ew)
         if debug:
-            print("layer ", i+1, "(after sparsification) edge_index size:", ei.shape)
+            print("layer ", i+1, "(after sparsification) edge_index size:", ei.shape, "homophily:", computeHomophily(data, ei))
     if nlayers > len(medge_index):
         print("Use less number of layers for the given", eps, " threshold, maximum:", len(medge_index), "layers")
         sys.exit(1)
@@ -151,6 +156,9 @@ def markov_process_disj_sparse(data, eps, inflate, nlayers, row_normalization =
         remaining_edge_idx = torch.nonzero(ew >= eps).flatten()
         ei = ei[:,remaining_edge_idx]
         ew = ew[remaining_edge_idx]
+        if ei.shape[1] < 1:
+            print("No more edges..! stopping after ", i, "layers")
+            break
         # normalization
         if row_normalization:
             ei, ew = markov_normalization(ei, ew, len(data.x), 'row')

utils.py

@@ -6,6 +6,8 @@
 import torch_geometric.transforms as T
 from torch_geometric.data import Data
 import torch
+from networkx.algorithms import community
+import networkx as nx
 
 def readInput3f(inputf, labelf, featuref, oneIndexed = False, onelabeled = False, debug = True):
     # inputf: input file name with path
@@ -178,3 +180,35 @@ def loadPyGDataset(dataset_name = 'Cora'):
     dataset = Planetoid(path, dataset_name, num_train_per_class=30, transform=T.NormalizeFeatures())
     data = dataset[0]
     return data
+
+# compute homophily
+def computeHomophily(data, ei = None):
+    if ei is None:
+        edges = data.edge_index.t()
+    else:
+        edges = ei.t()
+    nominator = 0
+    for edge in edges:
+        nominator += data.y[edge[0]] == data.y[edge[1]]
+    return nominator / len(edges)
+
+# compute community mixing
+def mixingCommunityScore(data):
+    G = nx.Graph(data.edge_index.t().tolist())
+    comm = community.greedy_modularity_communities(G)
+    gd = dict()
+    for com in range(len(comm)):
+        for node in list(comm[com]):
+            gd[node] = com
+    count = 0
+    for edge in data.edge_index.t():
+        count += gd[edge[0].item()] == gd[edge[1].item()]
+    return count / len(data.edge_index.t())
+
+# compute new edges percentage
+def newEdges(data, edges):
+    count = 0
+    for edge in edges.t():
+        if edge not in data.edge_index.t():
+            count += 1
+    return 100.0 * count / len(edges.t())