kangchenghou / gntk Goto Github PK

clf = GridSearchCV(svc, {'C' : C_list}, cv=zip(train_fold_idx, test_fold_idx), n_jobs=80, verbose=0, return_train_score=True)

I think you need something similar for return_train_score=True on line:
clf = GridSearchCV(svc, {'C' : C_list}, cv=zip(train_fold_idx, test_fold_idx), n_jobs=80, verbose=0)?

       `            if node_features != []:
            node_features = np.stack(node_features)
            node_feature_flag = True
        else:
            node_features = None
            node_feature_flag = False

        assert len(g) == n

        g_list.append(S2VGraph(g, l, node_tags))`

So it looks like node_features is never really used, and node_tags is used. However, none of the other files gram.py, gntk.py, search.py etc use node_tags. And gntk uses node_features, which matches the paper, but its never set here.

Could you please let me know, how this method can be used on my own data-set. How can I format my data-set to fit into your model ?