def label_evaluation(test_data, predicted_labels):
gold_labels = flatten([flatten(i["gold_labels"]) for i in test_data])
gold_labels = [1 if i else 0 for i in gold_labels]
metric_evaluation = Evaluator()
metric_evaluation.compute_all(gold_labels, predicted_labels)
log.write("Confusion Matrix :")
log.write(metric_evaluation.confusion_matrix)
log.write("Accuracy = %f" % metric_evaluation.accuracy)
log.write("Precision = %f" % metric_evaluation.precision)
log.write("Recall = %f" % metric_evaluation.recall)
log.write("F1 Score = %f" % metric_evaluation.f1_score)
