运营商数据分割成多ple tables etc

aldavenport

Hello,
I am working on a college project of which involves taking a CSV file and breaking the data by class into different tables. My goal is to then run a probability assessment against these tables and hopefully output a percentage. I am new to RapidMiner and am having some issues with the operators. Does anyone here have any helpful tips? Anything would be greatly appreciated.

rfuentealba

Hello@aldavenport,

If your CSV has something like:

id,att1,att2,class
0,0,0,none
1,1,1,odd
2,2,1,even
3,3,2,odd
4,4,3,even
5,5,5,odd
6,6,8,even

There are many ways to do this. The simple one would be a chain of "Filter Examples" plus "Store". YouRead CSVand then apply as manyFilter Examplesoperators as classes you have, and then store these (wherever you want)



A more complex one would be this:



This one multiplies the data from the CSV. First line (on the top) it selects and removes the duplicates of your data, and then it loops over each information related with the class.

(Make sure this is the order you have at this level)


Inside theLoop Examplesoperator, you can have this:



That uses theExtract Macrooperator to filter the examples recalled and store these, no matter how many classes you have.

Hope this helps,

Rodrigo.

aldavenport

Thank you Rodrigo! I am going to give these examples a shot. Again, thank you!

MartinLiebig

Another nice option is the Group Into Collection operator from Operator Toolbox.

Best,
Martin

Chemical_eng

Hello I have tried the approach of the loop , but instead of getting separate tables per category, I am getting a table with the unique categories, what am I doing wrong ? thanks