df.mensual <- df %>%
group_by(anio,mes,codigo) %>%
summarise(precip.mensual = sum(lluvia,na.rm="TRUE"))

filtering = filter(df.mensual, mes<3)