Importing data as Binominal

ppsheehy

Hi,

I’m pretty new to RapidMiner, so hopefully this isn’t  stupid question….

I am trying to use FP Growth to identify frequent itemsets. The data has about 900 attributes and there are about 5500 rows. The data is either true or false. I initially tried importing this as CSV and then connecting to an FP-Growth operator. However, I get an error saying the FP-Growth requires Binominal data. So I added a Nominal2Binominal operator. However, then I get an Out of Memory error. I also tried using a 'Read from AML' operator and then editing the attribute file changing Nominal to Binominal, but that didn’t work…

My question is: Is there a way to get the data to be imported or read as Binominal in the first place? It alsways seems to default to Nominal.

Any help appreciated,

Thanks

Paul

land

Hi Paul,
at first: The upcoming 5.1 version will have greatly extended Import Wizards, that will offer to import the data correctly in the first place.

Till then, I would recommend to update to the current 5.0.007 version of RapidMiner and then include a process like the one that can be retrieved from myExperiment.org via the Community Extension. The process is called "Preprocessing nominal data for frequent item set mining". It shows how to preprocess your data after importing.

Greetings,
  Sebastian

ppsheehy

Hi Sebastian,

Thanks a million for that - I would not have thought of that way of changing the inputs to Binominal. Very neat. Thanks also for cluing me in to myExperiment. I had not really understood what was available there before.

Again many thanks for the solution. Looking forward to 5.1.....

Paul