topic Re: What am I doing Wrong in SAS Programming

What am I doing Wrong

timmelvin981 — Mon, 23 Apr 2018 23:30:49 GMT

I am trying to run a simple Anova procedure. Could anyone tell me by looking at the below code what am I doing wrong? I need to get some type of output data beside just listing the dollar amount to the number next to it.

DATA ONE; FIRSTOBS=2 DLM='09'X
INPUT FREQ $ LOC;
Cards;
$36,198 1
$44,166 4
$31,874 4
$32,691 1
$41,259 4
$32,126 4
$42,924 2
$39,338 2
$35,166 1
$33,664 1
$41,027 4
$31,159 4
$37,166 3
$34,696 3
$33,226 3
$33,386 3
$35,166 1
$38,655 1
$31,835 2
$43,235 2
$40,600 2
$35,901 3
$34,505 3
$31,184 1
$30,064 3
$27,274 4
$30,844 3
$35,358 4
$34,280 2
$48,631 2
$31,960 4
$43,839 2
$30,778 1
$32,019 3
$33,096 3
$31,606 3
$33,549 4
$41,901 2
$39,196 2
$32,306 1
$29,851 3
$34,098 1
$38,091 3
$33,081 4
$35,541 2
$37,848 1
$36,335 4
$32,533 1
$33,546 3
$43,269 4

INPUT FREQ LOC;
IF LOC=1 THEN REGION='Southeast';
IF LOC=2 THEN REGION='Northeast';
IF LOC=3 THEN REGION='Midwest';
IF LOC=4 THEN REGION='WEST';
RUN;
PROC PRINT;
RUN;
PROC ANOVA;
CLASS REGION;
MODEL MEDINC=REGION;
MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.05;
RUN;
*/PROC GLM;
*CLASS REGION;
*MODEL FREQ=REGION;
*CONTRAST 'CENTRAL CITY VS. SUBURBS AND RURAL' REGION -1 0.5 0.5;
*MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.1;
*RUN;
/*PROC NPAR1WAY ANOVA WILCOXON;
CLASS REGION;
VAR FREQ;
RUN;*/

Re: What am I doing Wrong

Astounding — Tue, 24 Apr 2018 00:40:03 GMT

First, let's clean up the DATA step in a few ways. Abbreviating the data lines for readability (you wouldn't actually do that), here is an improved version:

DATA ONE;

informat FREQ dollar8.;
INPUT FREQ LOC;

IF LOC=1 THEN REGION='Southeast';
else IF LOC=2 THEN REGION='Northeast';
else IF LOC=3 THEN REGION='Midwest';
else IF LOC=4 THEN REGION='WEST';
Cards;
$38,655 1
$31,835 2
$43,235 2
$40,600 2
$35,901 3
$34,505 3

;

This version makes FREQ a numeric variable instead of character.

In the ANOVA, one thing stands out as possibly being a problem. In a MEANS statement, are you permitted to use a character variable? (I actually don't know the answer to that.) Instead of using REGION, you could certainly use LOC (in all three places).

PROC ANOVA;
CLASS REGION;
MODEL MEDINC=REGION;
MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.05;
RUN;

Re: What am I doing Wrong

timmelvin981 — Tue, 24 Apr 2018 01:48:10 GMT

Thanks for the help but it is still not wanting to run the ANOVA

Re: What am I doing Wrong

Astounding — Tue, 24 Apr 2018 02:07:11 GMT

You'll have to help out here. Post the log, so we can see what is not working.

Re: What am I doing Wrong

art297 — Tue, 24 Apr 2018 02:08:17 GMT

The following works for me:

DATA ONE; 
  infile cards DLM='09'X;
  INPUT medinc comma7. LOC;
  IF LOC=1 THEN REGION='Southeast';
  else IF LOC=2 THEN REGION='Northeast';
  else IF LOC=3 THEN REGION='Midwest';
  else IF LOC=4 THEN REGION='WEST';
  Cards;
$36,198 1
$44,166 4
$31,874 4
$32,691 1
$41,259 4
$32,126 4
$42,924 2
$39,338 2
$35,166 1
$33,664 1
$41,027 4
$31,159 4
$37,166 3
$34,696 3
$33,226 3
$33,386 3
$35,166 1
$38,655 1
$31,835 2
$43,235 2
$40,600 2
$35,901 3
$34,505 3
$31,184 1
$30,064 3
$27,274 4
$30,844 3
$35,358 4
$34,280 2
$48,631 2
$31,960 4
$43,839 2
$30,778 1
$32,019 3
$33,096 3
$31,606 3
$33,549 4
$41,901 2
$39,196 2
$32,306 1
$29,851 3
$34,098 1
$38,091 3
$33,081 4
$35,541 2
$37,848 1
$36,335 4
$32,533 1
$33,546 3
$43,269 4
;

PROC ANOVA;
CLASS REGION;
MODEL MEDINC=REGION;
MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.05;
RUN;

Art, CEO, AnalystFinder.com

Re: What am I doing Wrong

timmelvin981 — Tue, 24 Apr 2018 02:36:52 GMT

Thank you! The one that you did worked for me also. I am also attempting to do this code with the same one.

/PROC GLM;
CLASS REGION;
MODEL FREQ=REGION;
CONTRAST 'CENTRAL CITY VS. SUBURBS AND RURAL' REGION -1 0.5 0.5;
MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.1;
RUN;
PROC NPAR1WAY ANOVA WILCOXON;
CLASS REGION;
VAR FREQ;
RUN;

It is also not working. I really appreciate your help as I am beyond a SAS novice!

Re: What am I doing Wrong

art297 — Tue, 24 Apr 2018 04:00:59 GMT

Like with your original code, you refer to a variable called FREQ, but don't have any such variable in your dataset. I presume that you were trying to run:

PROC GLM data=one;;
CLASS REGION;
MODEL MEDINC=REGION;
CONTRAST 'CENTRAL CITY VS. SUBURBS AND RURAL' REGION -1 0.5 0.5;
MEANS REGION/SCHEFFE TUKEY LSD SNK ALPHA=0.1;
RUN;

PROC NPAR1WAY data=one ANOVA WILCOXON;
CLASS REGION;
VAR MEDINC;
RUN;

Art, CEO, AnalystFinder.com

Re: What am I doing Wrong

timmelvin981 — Tue, 24 Apr 2018 13:19:36 GMT

Art,

Just wanted to say thank you for your help. It was very much appreciated.