def run_ICA(classifier_name, classifier_args, num_iter):
features, labels, train_, val_, test_, graph, domain_labels = load_data()
# run training
t_begin = time.time()
# random ordering
np.random.shuffle(val_)
y_true = [graph.Graph.nodes[node]['label'] for node in test_]
local_clf = LocalClassifier(classifier_name, classifier_args)
agg = pick_aggregator('count', domain_labels)
relational_clf = RelationalClassifier(classifier_name, agg,
classifier_args)
ica = ICA(local_clf, relational_clf, True, max_iteration=num_iter)
ica.fit(graph.Graph, train_)
print('Model fitting done...')
conditional_node_to_label_map = create_map(graph.Graph, train_)
ica_predict = ica.predict(graph.Graph, val_, test_,
conditional_node_to_label_map)
print('Model prediction done...')
ica_accuracy = accuracy_score(y_true, ica_predict)
t_end = time.time()
print(classification_report(y_true, ica_predict))
print(ica_accuracy)
elapsed_time = t_end - t_begin
print('Start time: \t\t' + time.strftime("%H:%M:%S", time.gmtime(t_begin)))
print('Elapsed time: \t\t' +
time.strftime("%H:%M:%S", time.gmtime(elapsed_time)))
print('End time: \t\t' + time.strftime("%H:%M:%S", time.gmtime(t_end)))
args = parser.parse_args()
graph, domain_labels = load_linqs_data(args.content_file, args.cites_file)
kf = KFold(n=len(graph.node_list),
n_folds=args.num_folds,
shuffle=True,
random_state=42)
accuracies = []
cm = None
for train, test in kf:
clf = LocalClassifier(args.classifier)
clf.fit(graph, train)
y_pred = clf.predict(graph, test)
y_true = [graph.node_list[t].label for t in test]
accuracies.append(accuracy_score(y_true, y_pred))
if cm is None:
cm = confusion_matrix(y_true, y_pred, labels=domain_labels)
else:
cm += confusion_matrix(y_true, y_pred, labels=domain_labels)
print accuracies
print "Mean accuracy: %0.4f +- %0.4f" % (np.mean(accuracies),
np.std(accuracies))
print cm
conditional_map = {}
n = range(len(graph.node_list))
if not args.dont_evaluate_local:
local_accuracies = defaultdict(list)
relational_accuracies = defaultdict(list)
for t in range(args.num_trials):
for b in budget:
train, test = train_test_split(n, train_size=b, random_state=t)
# True labels
y_true = [graph.node_list[t].label for t in test]
if not args.dont_evaluate_local:
# local classifier fit and test
local_clf = LocalClassifier(args.classifier)
local_clf.fit(graph, train)
local_y_pred = local_clf.predict(graph, test)
local_accuracy = accuracy_score(y_true, local_y_pred)
local_accuracies[b].append(local_accuracy)
# relational classifier fit and test
agg = pick_aggregator(args.aggregate, domain_labels, args.directed)
relational_clf = RelationalClassifier(
args.classifier, agg, not args.dont_use_node_attributes)
relational_clf.fit(graph, train)
conditional_node_to_label_map = create_map(graph, train)
relational_y_pred = relational_clf.predict(
graph, test, conditional_node_to_label_map)
relational_accuracy = accuracy_score(y_true, relational_y_pred)
relational_accuracies[b].append(relational_accuracy)
parser.add_argument('-classifier', default='sklearn.linear_model.LogisticRegression', help='The underlying classifier.')
parser.add_argument('-num_folds', type=int, default=10, help='The number of folds.')
args = parser.parse_args()
graph, domain_labels = load_linqs_data(args.content_file, args.cites_file)
kf = KFold(n=len(graph.node_list), n_folds=args.num_folds, shuffle=True, random_state=42)
accuracies = []
cm = None
for train, test in kf:
clf = LocalClassifier(args.classifier)
clf.fit(graph, train)
y_pred = clf.predict(graph, test)
y_true = [graph.node_list[t].label for t in test]
accuracies.append(accuracy_score(y_true, y_pred))
if cm is None:
cm = confusion_matrix(y_true, y_pred, labels = list(domain_labels))
else:
cm += confusion_matrix(y_true, y_pred, labels = list(domain_labels))
print accuracies
print "Mean accuracy: %0.4f +- %0.4f" % (np.mean(accuracies), np.std(accuracies))
print cm
test = idx_val
else:
test = idx_test
eval_idx = np.setdiff1d(range(adj.shape[0]), idx_train)
# run training
ica_accuracies = list()
for run in range(args.num_trials):
t_begin = time.time()
# random ordering
np.random.shuffle(eval_idx)
y_true = [graph.node_list[t].label for t in test]
local_clf = LocalClassifier(args.classifier)
agg = pick_aggregator(args.aggregate, domain_labels)
relational_clf = RelationalClassifier(args.classifier, agg)
ica = ICA(local_clf,
relational_clf,
args.bootstrap,
max_iteration=args.max_iteration)
ica.fit(graph, train)
conditional_node_to_label_map = create_map(graph, train)
ica_predict = ica.predict(graph, eval_idx, test,
conditional_node_to_label_map)
ica_accuracy = accuracy_score(y_true, ica_predict)
ica_accuracies.append(ica_accuracy)
print 'Run ' + str(run) + ': \t\t' + str(
ica_accuracy) + ', Elapsed time: \t\t' + str(time.time() - t_begin)
n=range(len(graph.node_list))
if not args.dont_evaluate_local:
local_accuracies = defaultdict(list)
relational_accuracies = defaultdict(list)
for t in range(args.num_trials):
for b in budget:
train, test = train_test_split(n, train_size=b, random_state=t)
# True labels
y_true=[graph.node_list[t].label for t in test]
if not args.dont_evaluate_local:
# local classifier fit and test
local_clf=LocalClassifier(args.classifier)
local_clf.fit(graph,train)
local_y_pred=local_clf.predict(graph,test)
local_accuracy=accuracy_score(y_true, local_y_pred)
local_accuracies[b].append(local_accuracy)
# relational classifier fit and test
agg=pick_aggregator(args.aggregate,domain_labels,args.directed)
relational_clf=RelationalClassifier(args.classifier, agg, not args.dont_use_node_attributes)
relational_clf.fit(graph,train)
conditional_node_to_label_map=create_map(graph,train)
relational_y_pred=relational_clf.predict(graph,test,conditional_node_to_label_map)
relational_accuracy=accuracy_score(y_true,relational_y_pred)
relational_accuracies[b].append(relational_accuracy)
