yy = []
pp = []
seq_pred_y_valid = np.zeros((_seq_pred_y_valid.shape[0], before_steps))
seq_pred_y_valid[np.where(
_seq_pred_y_valid[:, :before_steps] > sur_thrld_valid)] = 1
early_correct = np.sum(seq_pred_y_valid == batch_y_valid, axis=0)
yy.extend(batch_y_valid)
pp.extend(seq_pred_y_valid)
yy = np.asarray(yy)
pp = np.asarray(pp)
gt = np.logical_not(yy).astype(float)
pr = np.logical_not(pp).astype(float)
_precision, _recall, _F1 = prec_reca_F1(gt, pr)
seq_corr_rate = early_correct / float(_seq_pred_y_valid.shape[0])
thrld_score[sur_thrld_valid] = np.mean(seq_corr_rate)
best_thrld, acc_best_threshold = max(thrld_score.iteritems(),
key=operator.itemgetter(1))
# print best_thrld, acc_best_threshold, seq_corr_rate
# # exit(0)
# best_thrld = 0.9505
#
# print "-----", best_thrld
yy = []
pp = []
x = [[0, 1, 1, 0], [1, 1, 1, 0], [0, 1, 1, 1], [0, 0, 1, 0], [0, 0, 1, 0]] y = [[1, 0, 1, 0], [0, 0, 0, 0], [0, 1, 1, 1], [1, 0, 1, 0], [1, 1, 0, 1]] x = np.array(x) y = np.array(y) print x print print print print y precision, recall, F1 = prec_reca_F1(x, y) print print print "precison: ", precision print print "recall: ", recall print print "F1: ", F1 print exit(0)