topic Re: BALANCED SAMPLE in SAS Enterprise Guide

BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 18:51:47 GMT

How can I do a balanced sample: example

I have a data with 1000 clients, being of these, 300 are not good and 700 are good (binary, independent variable), and I need to do a proporcional balanced sample, as if I divide 50% good and 50% bad...somebody can help me

Re: BALANCED SAMPLE

PaigeMiller — Tue, 30 Jan 2018 18:53:52 GMT

Randomly select X from the not good, and X from the good.

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 18:56:27 GMT

How can I do this in SAS Enterprise Guide by proc, can you help me?

Re: BALANCED SAMPLE

Reeza — Tue, 30 Jan 2018 19:00:29 GMT

Why not set priors in your regression instead?

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 19:02:36 GMT

I did this in SAS Enterprise Miner...but my sample was not good

Re: BALANCED SAMPLE

Reeza — Tue, 30 Jan 2018 19:08:15 GMT

I don't see a task that would do this in SAS EG 7.12

I think you need a code node and PROC SURVEYSELECT

You could possibly do this in a query by first calculating the numbers needed and then using a random number generated to select that many from each group.

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 19:08:31 GMT

I did this in SAS Enterprise Miner...but my sample was not good, some suggestions?

Re: BALANCED SAMPLE

Reeza — Tue, 30 Jan 2018 19:09:30 GMT

my sample was not good

Your sample or prediction wasn't 'good'? What do you mean by that?

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 19:10:13 GMT

that's ok I try. tks

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 19:11:16 GMT

my prediction was not good

Re: BALANCED SAMPLE

PRISCILABRA — Tue, 30 Jan 2018 19:17:16 GMT

my prediction, ROC was very low...I need to do a random before by proc...understand? I try to do with proc surveyselect, thanks a lot. bye

Re: BALANCED SAMPLE

Reeza — Tue, 30 Jan 2018 19:24:33 GMT

Actually, you could take all of the 30% sample and then some of the other sample until the N's match, that's should be relatively easy in a data step.

*create sample data - with  a random number;
data have;
	call streaminit(25);

	*set stream for random number - means same results every time;
	group=1;

	do i=1 to 30;
		x=rand('normal', 25, 5);
		output;
	end;

	group=2;

	do i=1 to 70;
		x=rand('normal', 35, 2);
		output;
	end;
run;

*sort with random number to randomize;
proc sort data=have;
	by group x;
run;

data want;
	set have;
	by group;
	retain counter flag max_count;

	if _n_=1 then
		flag=0;

	if first.group then
		counter=0;
	counter+1;

	if last.group and flag=0 then
		do;
			flag=1;
			max_count=counter;
		end;

	if flag=1 and counter> max_count then
		stop;;
run;

Re: BALANCED SAMPLE

PRISCILABRA — Wed, 31 Jan 2018 11:00:24 GMT

Goog Morning,

I tried and it worked out. Thanks a lot

Re: BALANCED SAMPLE

Ksharp — Wed, 31 Jan 2018 13:26:55 GMT

Are you using German Credit.xlsx for making a CreditCard ?

proc import datafile="/courses/d8fb3215ba27fe300/1--German Credit.xlsx" 
out=have dbms=xlsx replace;
run;
proc sort data=have;by good_bad;run;
proc surveyselect data=have out=want sampsize=(300 300) seed=12345678;
strata good_bad;
run;
proc freq data=want;
table good_bad;
run;