DSXchange

Add a new column to your input. This column has the same value, say "X", in every row. A Column Generator stage would be ideal.

Create a Data Set containing pairs of values, "X" and a value, for example X,100; X,125; X,150 and so on up to the largest possible value. Permit duplicates.

Effect a join between this data set and your input stream.

Filter the result on Amt <= Value.

There could be more simpler & smarter solutions suggested further .
One of the naive possibilities include 2-3 jobs.

Counter Initialization based on keys could be 1 DS job or could be merged.
Subsequently,have a single DS job which does this validation , sorted on ascending counter for the keys.
Have this DS Job created in step 2 be executed through sequencer based on the check whether the Indicator is N for any records sorted on Keys and counter.

Partitioning will still be on keys and counter, with same sorting option

Hoping this helps.

Maybe... but Ray's solution is pretty darn simple already.

Mike

Or build an operator