# Macro averaged in binary classification

Is macro averaged the precision, recall and F1 score recommended for binary classification data? The calculation of precision is calculated for both class and averaging it makes it seems better. Or macro averaged is works better for multiclass classification only?