import sys import os import fileio import gspan if __name__ == '__main__': print 'Database: ', sys.argv[1] database = fileio.read_file(sys.argv[1]) print 'Number Graphs Read: ', len(database) print 'Support: ', sys.argv[2], minsup = int((float(sys.argv[2]) * len(database))) print minsup database, freq, trimmed, flabels = gspan.trim_infrequent_nodes( database, minsup) database = fileio.read_file(sys.argv[1], frequent=freq) print 'Trimmed ', len(trimmed), ' labels from the database' print flabels gspan.project(database, freq, minsup, flabels)
else: cons = (ml_cons, cl_cons[: num_constraints - len_ml_cons]) for m in model: for length in [5]: scores = {names[0]: [], names[1]: [], names[2]: []} times = [] for k in xrange(0, k_fold): train_file = output_file_train + str(k) + ".txt" test_file = output_file_test + str(k) + ".txt" database_train = fileio.read_file(train_file) print "Number Graphs Read: ", len(database_train) minsup = int((float(min_sup) * len(database_train))) print minsup database_train, freq, trimmed, flabels = gspan.trim_infrequent_nodes(database_train, minsup) database_train = fileio.read_file(train_file, frequent=freq) train_labels = np.array(graph_labels_train[k]) tik = datetime.utcnow() pattern_set_global = [] class_index = 1 # for class_index in xrange(num_classes): H, L, L_hat, n_graphs, n_pos, n_neg, pos_index, neg_index, graph_id_to_list_id = fileio.preproscessing( database_train, class_index, labels_mapping, m ) X_train, pattern_set_global = gspan.project( database_train, freq, minsup, flabels,
def Gspan(support): database = fileio.read_file(r"database.txt") minsup = int((float(support)*len(database))) database, freq, trimmed, flabels = gspan.trim_infrequent_nodes(database, minsup) database = fileio.read_file(r"database.txt", frequent = freq) gspan.project(database, freq, minsup, flabels)