Replacing a Date for a New Date

mmagnuson — Mon, 02 Oct 2017 15:01:54 GMT

Hello,

I have records that are populated for each month of the calendar year that have a column for points. Points change per month. There is also one record per month. For example, John Smith has one record for Jan, one record for Feb, etc. I pull this report every month so the reporting month changes.

I want to freeze his points on a specific month without impacting the other people in the dataset. So if I want to freeze his results as of June then I would not want to see July, Aug, Spet, or October on the report. As of right now I used an excel doc to import names of people I wanted to do this for and have excluded months that I don't want (these months are indicated on the excel sheet). I.e. in the excel sheet i have John Smith and July 1, 2017. So I excluded based on the date but I'm unsure of how to REPLACE the JUNE month with September so that September comes through with populated data of July. I tried to use a prompt but am unsure of how to go about doing this.

Hope this makes sense!

Re: Replacing a Date for a New Date

Kurt_Bremser — Mon, 02 Oct 2017 15:38:03 GMT

Suppose dataset have has id, date, and points. Dataset updates has id and date.

Sort both datasets by id and date, then run

data want;
merge
  have (in=a)
  updates (in=b)
;
by id date;
retain save_points;
if a;
if first.id then save_points = .;
if b then save_points = points;
if save_points ne . then points = save_points;
drop save_points;
run;

topic Replacing a Date for a New Date in SAS Enterprise Guide

Replacing a Date for a New Date

Re: Replacing a Date for a New Date