topic Re: Dividing a data from single column into groups of 20. in SAS Programming

Dividing a data from single column into groups of 20.

Discaboota — Thu, 27 Oct 2022 12:28:09 GMT

Hey guys, I am new to SAS and I have learned a lot from this community.
I want to divide the data of 4.5 million from a column into groups of 20, doesn't matter about the number of groups but the data keeps on increasing every day by around 10k so I need to code it so that my code can cope with increasing data.
Please Help. Thank you.

Re: Dividing a data from single column into groups of 20.

PeterClemmensen — Thu, 27 Oct 2022 12:30:08 GMT

Groups of 20 what?

Also, why do you want to split up your data? SAS can easily handle a data set of 4.5M obs.

Re: Dividing a data from single column into groups of 20.

PaigeMiller — Thu, 27 Oct 2022 12:46:39 GMT

Why? How does dividing data into groups of 20, however you define these groups, help anything?

Re: Dividing a data from single column into groups of 20.

Tom — Thu, 27 Oct 2022 13:31:33 GMT

Not sure what you mean by divide? Do you want to assign a GROUP variable? Does it need to be random assignment?

If not then simply use their position in the dataset.

data want;
  set have;
  group = ceil(_n_/20);
run;

Re: Dividing a data from single column into groups of 20.

Discaboota — Fri, 28 Oct 2022 09:20:26 GMT

Hey, I am required to provide unique leads of 20 customers every day to every sales officer my bank has, so I was thinking of dividing it into groups of 20 and then providing the leads to them sequentially.
If I can do it in a better way then please let me know.

Re: Dividing a data from single column into groups of 20.

Discaboota — Fri, 28 Oct 2022 09:23:58 GMT

Thank you Tom for your help, it is what I was looking for. I am thinking of dividing the data into groups so that I can deliver groups sequentially to the required person.

Re: Dividing a data from single column into groups of 20.

ballardw — Fri, 28 Oct 2022 15:53:13 GMT

@Discaboota wrote:
Thank you Tom for your help, it is what I was looking for. I am thinking of dividing the data into groups so that I can deliver groups sequentially to the required person.

If the original sequence is needed then perhaps the below. This increments a counter every 20 observations.

data want;
  set have;
  retain group 1;
  if mod(_n_,20)=0 then group+1;
run;