That was a nice solution. The other option would be to split the data frame into two parts and fix the names in one part and then join them.