How to remove column with low counts

Hi all,

I need some help!  The data have many columns and each column contains 0 or 1s.  I would like to remove the columns with smaller number of 1s.  Please help!




you can use a Column Aggregator to compute the sum of all valus per row and then just Row Filter based on that sum.

I think in this case GroupBy or just Statistics node should be used. (The task is to remove columns, not rows.) After that I am afraid Xiaojie have to remove the columns manually based on those statistics.

Oh.. right... stupid me.

Let me redeem myself by remarking that there's no need for manual work. Two Transpose's alone should suffice, although they might be too expensive. Combine them with an Extract Column Header and a Reference Column Filter though, and the extra cost can be reduced to almost none.

1 Like

Once you have a list of columns you can probably format it in such a way you can use a reference row or column filter