Select first n attributes from example set

Dear all,
Is there an efficient way to select first n attributes from an example set? In my process I have reordered the attributes in ascending way. Now I want to proceed with just the first 5 attributes.
I was thinking of several combinations of "select attributes", "work ok subset", "branch", "loop attributes" but none of them worked.
Looking forward to any comments!
Best regards
Sachs
Best Answer
-
Wonder why I haven't seen this operator before but "remove attribute range" is the operator that solves my problem! It can be combined with "extract macro" to count the number of attributes first.
1
Answers
-
Hi,
i would go for: Filter Examples (remove all), Transpose, Filter Examples Range (Select the atts you want), Transpose, Data to Weights. The result is a weight vector you can apply with Select by Weights to your original data set.
~Martin
0 -
Hi Martin,
Thanks for your help. That's indeed a good approach that could work but unfortunately not in my case. My example set has many many many attributes and just a few examples. Once I transpose the example set I exceed my license of 10.000 examples as there are even more attributes...
Best regards
Sachs
0 -
Wonder why I haven't seen this operator before but "remove attribute range" is the operator that solves my problem! It can be combined with "extract macro" to count the number of attributes first.
1