On Feb 7, 12:29 pm, Walt <walt_ask..._at_SHOESyahoo.com> wrote:
> Walt wrote:
> Ok, here's the deal. Trimming according to cume_dist or percent_ranking
> trims by value, not by record. That is, all the records with n=10 come
> and go as a group - either they're in the sample used to compute the
> average or they're not.
>
> John's first formula trims the top and bottom 20% of records, the others
> trim all records with a value between the 20th percentile and 80th
> percentile. Not the same thing
Good points all. OK, then go with the version based on ROWNUM.
I see you caught the issue w/ the boundary condition. I should've
used a better test case!
Received on Wed Feb 07 2007 - 11:43:01 CST
