topic How do I correctly recode these variables in SAS Programming

How do I correctly recode these variables

sc123 — Fri, 13 Sep 2024 20:06:23 GMT

Create the following new variables:

SHOCK: 0=no shock, 1=shock (based on the variable shoktype)

DIED: 0=lived, 1=died (based on the variable survive)

AGECAT: (1=<45, 2=45-55, 3=56-65, 4=>65).

The INPUT is

INPUT

#1 IDNUM 1-4 AGE 5-8 SEX 13-15 SURVIVE 16 SHOKTYPE 17-20 SBP1 21-24

MAP1 25-28 HEART1 29-32 CARDIAC1 45-48 2 URINE1 57-60 HGB1 69-72 1

#2 SBP2 21-24 MAP2 25-28 HEART2 29-32 CARDIAC2 45-48 2 URINE2 57-60 HGB2 69-72 1;

 *First always run a descriptive stat to ensure there are no missing values, there were no missing values for this dataset;
 PROC FORMAT; *Recommended to include PROC FORMAT in the very beginning of the code;
 VALUE SEXFMT 1= "Male"
           2= "Female";
           
 VALUE SHOCKFMT 2= "Non-Shock"
             3= "Hypovolemic Shock"
             4= "Cardiogenic Shock"
             5= "Bacterial Shock"
             6= "Neurogenic Shock"
             7= "Other"; *CHANGE??;
             
 VALUE DIEDFMT  1= "Lived"
                3="Died";
           
 VALUE AGECATFMT 1="<45"
                 2 = "45-55"
                 3 = "56-65"
                 4 = ">65";
                 
RUN;
DATA AFIFI;
INFILE '/home/u63916728/My Sas File/Class Raw Data Files sasuser.v94/AFIFI-2.DAT';
INPUT 
#1 IDNUM 1-4 AGE 5-8 SEX 13-15 SURVIVE 16 SHOKTYPE 17-20 SBP1 21-24 

       MAP1 25-28 HEART1 29-32 CARDIAC1 45-48 2 URINE1 57-60 HGB1 69-72 1 

#2 SBP2 21-24 MAP2 25-28 HEART2 29-32 CARDIAC2 45-48 2 URINE2 57-60 HGB2 69-72 1; 

LABEL 
MAP1 = 'Mean Arterial Pressure 1 (mm Hg)'
HEART1 = 'Heartrate 1 (Beats Per Minute)'
CARDIAC1 = 'Cardiac Index 1'
MAP2 = 'Mean Arterial Pressure 2 (mm Hg)'
HEART2 = 'Heartrate 2 (Beats Per Minute)'
CARDIAC2 = 'Cardiac Index 2'; *End here;
*Label was created in the DATA Step;

*Now we will recode variables;
if shoktype <=0 and shoktype ne . then shock=0;
else if shoktype>=1 then shock=1;

if survive<=0 and survive ne . then died=0;
else if survive>0 then died=1;

IF AGE NE . AND AGE<45 THEN AGECAT=1;
ELSE IF 45<=AGE<56 THEN AGECAT=2;
ELSE IF 56<=AGE<66 THEN AGECAT =3;
ELSE IF AGE >=66 THEN AGECAT = 4; *MIGHT BE ABLE TO DO >65;

FORMAT SEX SEXFMT. SHOKTYPE SHOCKFMT. SURVIVE DIEDFMT. AGE AGECATFMT.; 
RUN;

Below is what I have tried. Agecat is the only proper recoded variable.

Re: How do I correctly recode these variables

ballardw — Fri, 13 Sep 2024 21:24:27 GMT

I don't see any example input values for Shoktype, Survive as read by the input statement.

It does look like the format Shockfmt should be used with variable Shock not with Shoktype.

Similar the format Diedfmt looks like it should be with Died, not Survive.

FORMAT SEX SEXFMT. SHOKTYPE SHOCKFMT. SURVIVE DIEDFMT. AGE AGECATFMT.;

For 99 percent of the time I would not have recoded age at all, just define the format for age categories as

VALUE AGECATFMT 
0 -<45 ="<45"
45-55 = "45-55"
56-65 = "56-65"
66-high = ">65";

Assuming age is integer values. Just use that as the format for the Age variable for any analysis, reporting or graphing. My permanent format libraries have about a dozen different age category formats that are used as needed on the age value. That way I don't have to keep recoding age into 4 (or 6 or 23 or 8 groups).

Re: How do I correctly recode these variables

Tom — Sat, 14 Sep 2024 01:41:40 GMT

Wow. A data file with implied decimal points. I haven't seen one of those in over 35 years.

data test;
  input CARDIAC2 1-4 2  ;
cards;
1234
567
;

Result

Obs    CARDIAC2

 1       12.34
 2        5.67