RapidMiner 9.8 Beta is now available
Be one of the first to get your hands on the new features. More details and downloads here:
speeding up Numerical To Binomial?
The operator Numerical To Binomial does the trick... slowly. For my table of (test!) data (1.5mm rows of 6k columns) it is now taking 2.5 hours and no sign of it being complete.
Does anyone have a trick to speed things up?
One thing I thought of is writing the table to a temp CSV file and then use the smart guessing of the CSV import to make them binomial.
I've tried the GuesTypes operator, but that I killed after having run for an hour or so...
Writing to a tmp database table does not work as my postgresql database does not like more than 1600 columns.