(get list with) count of non-null values of each of multiple columns

[ooptennoort]

I have a data set with >100 cols, most of which contain only 1 value (useless). I need to remove these so I have cols left with >1 value (the good stuff).

So it seems to me I somehow need to List.Select all the List.NonNullCount >1 for each _column (which suggests using Table.ColumnNames) but I just can't figure out how to put these together, followed of course by Table.SelectColumns with that list (the easy bit)

Any ideas (chatGPT provided absolute crap)?

[ppm1 · Accepted solution]

If all of the empty columns show as "null", you can unpivot those columns and the null rows will not be kept. You should consider keeping it unpivoted, but you can pivot it back out if needed. If they con't show as "null", you can unpivot and then filter out all the ones w/o values (blank).

Pat

[ooptennoort]

So simple! Thanks!