how to create a code for likelihood ratio test?

[Nouf2]

Hello, 

I am taking a Biostatistics class and we use SAS. One homework question I couldn't figure out how to write the code

"Taking obesity as response variable, conduct model selection using backward method with maximum model including following covariates: WASIT_HIP_RATIO, Louisa, SBP, AGE, chol, stab_glub, glyhb, HDL, and MALE. " 

I followed the likelihood ratio test code that is provided on the course material note which is:

"proc logistic data=temp; 

 model obesity(event='1')=Waist_Hip_ratio Louisa SBP AGE chol stab_glub glyhb HDL MALE; 

 contrast 'lrt' Waist_Hip_ratio 1; Louisa 1; SBP 1; AGE 1; chol 1; stab_glub 1; glyhb 1; HDL 1; MALE 1;

 run;"

However, this code only work for two nested models according to notes!. 

My question: how to create a code for the quoted problem above? 

PS: I am a healthcare student and very basic with SAS. 

[PaigeMiller]

If the code you are showing produces errors or warnings, show us the LOG. We need to see the entire log from PROC LOGISTIC from the first line of code all the way down to the last NOTE after the PROC, all of it, 100% of the log for this PROC, with nothing chopped out or re-ordered. Paste the log as text into the window that appears when you click on the </> icon, this causes the log to be properly formatted (I no longer try to read logs that are not properly formatted).

If the output is not what you expect, show us the output. A screen capture of incorrect output is fine.

[Nouf2]

Thank you for your reply. I ran all the codes to give you a better understanding of what I want. 

 
 1          OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 72         
 73         PROC IMPORT OUT= WORK.temp DATAFILE= "/folders/myfolders/diabetes.csv"
 74         DBMS=CSV REPLACE;
 74       !                   GETNAMES=YES; DATAROW=2;
 75         RUN;
 
 NOTE: Unable to open parameter catalog: SASUSER.PARMS.PARMS.SLIST in update mode. Temporary parameter values will be saved to 
 WORK.PARMS.PARMS.SLIST.
 76          /**********************************************************************
 77          *   PRODUCT:   SAS
 78          *   VERSION:   9.4
 79          *   CREATOR:   External File Interface
 80          *   DATE:      21MAR21
 81          *   DESC:      Generated SAS Datastep Code
 82          *   TEMPLATE SOURCE:  (None Specified.)
 83          ***********************************************************************/
 84             data WORK.TEMP    ;
 85             %let _EFIERR_ = 0; /* set the ERROR detection macro variable */
 86             infile '/folders/myfolders/diabetes.csv' delimiter = ',' MISSOVER DSD lrecl=32767 firstobs=2 ;
 87                informat ID best32. ;
 88                informat CHOL best32. ;
 89                informat STAB_GLUB best32. ;
 90                informat HDL best32. ;
 91                informat CHOL_HDL_RATIO best32. ;
 92                informat GLYHB best32. ;
 93                informat LOCATION $10. ;
 94                informat AGE best32. ;
 95                informat GENDER $6. ;
 96                informat HEIGHT best32. ;
 97                informat WEIGHT best32. ;
 98                informat FRAME $6. ;
 99                informat SBP1 best32. ;
 100               informat DBP1 best32. ;
 101               informat SBP2 best32. ;
 102               informat DBP2 best32. ;
 103               informat WAIST best32. ;
 104               informat HIP best32. ;
 105               informat TIME best32. ;
 106               format ID best12. ;
 107               format CHOL best12. ;
 108               format STAB_GLUB best12. ;
 109               format HDL best12. ;
 110               format CHOL_HDL_RATIO best12. ;
 111               format GLYHB best12. ;
 112               format LOCATION $10. ;
 113               format AGE best12. ;
 114               format GENDER $6. ;
 115               format HEIGHT best12. ;
 116               format WEIGHT best12. ;
 117               format FRAME $6. ;
 118               format SBP1 best12. ;
 119               format DBP1 best12. ;
 120               format SBP2 best12. ;
 121               format DBP2 best12. ;
 122               format WAIST best12. ;
 123               format HIP best12. ;
 124               format TIME best12. ;
 125            input
 126                        ID
 127                        CHOL
 128                        STAB_GLUB
 129                        HDL
 130                        CHOL_HDL_RATIO
 131                        GLYHB
 132                        LOCATION  $
 133                        AGE
 134                        GENDER  $
 135                        HEIGHT
 136                        WEIGHT
 137                        FRAME  $
 138                        SBP1
 139                        DBP1
 140                        SBP2
 141                        DBP2
 142                        WAIST
 143                        HIP
 144                        TIME
 145            ;
 146            if _ERROR_ then call symputx('_EFIERR_',1);  /* set ERROR detection macro variable */
 147            run;
 
 NOTE: The infile '/folders/myfolders/diabetes.csv' is:
       Filename=/folders/myfolders/diabetes.csv,
       Owner Name=root,Group Name=vboxsf,
       Access Permission=-rwxrwx---,
       Last Modified=16Mar2021:21:49:57,
       File Size (bytes)=36704
 
 NOTE: 403 records were read from the infile '/folders/myfolders/diabetes.csv'.
       The minimum record length was 66.
       The maximum record length was 100.
 NOTE: The data set WORK.TEMP has 403 observations and 19 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.01 seconds
       cpu time            0.02 seconds
       
 
 403 rows created in WORK.TEMP from /folders/myfolders/diabetes.csv.
   
   
   
 NOTE: WORK.TEMP data set was successfully created.
 NOTE: The data set WORK.TEMP has 403 observations and 19 variables.
 NOTE: PROCEDURE IMPORT used (Total process time):
       real time           0.21 seconds
       cpu time            0.17 seconds
       
 
 148        data temp1; set temp;
 149        SBP= MEAN(SBP1,SBP2);
 150        Waist_Hip_ratio= waist/hip;
 151        if gender='male' then male=1; else if gender='female' then male=0;
 152        if location="Louisa" then Louisa=1; else if  location="Buckingham" then Louisa=0;
 153        BMI=weight*0.454/(height*0.0254)**2;
 154        if BMI>30 then OBESITY=1; else if BMI<=30 then OBESITY=0; if BMI=. then OBESITY=.;
 155        run;
 
 NOTE: Missing values were generated as a result of performing an operation on missing values.
       Each place is given by: (Number of times) at (Line):(Column).
       5 at 149:6    2 at 150:23   1 at 153:11   5 at 153:25   
 NOTE: There were 403 observations read from the data set WORK.TEMP.
 NOTE: The data set WORK.TEMP1 has 403 observations and 25 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.02 seconds
       cpu time            0.02 seconds
       
 
 156        
 157        proc logistic data=temp1;
 158         model OBESITY(event='1')= Louisa male Louisa*male;
 159         run;
 
 NOTE: PROC LOGISTIC is modeling the probability that OBESITY=1.
 NOTE: Convergence criterion (GCONV=1E-8) satisfied.
 NOTE: There were 403 observations read from the data set WORK.TEMP1.
 NOTE: PROCEDURE LOGISTIC used (Total process time):
       real time           0.32 seconds
       cpu time            0.29 seconds
       
 
 160        proc logistic data=temp1;
 161        class louisa male;
 162        model obesity(event='1')=louisa male;
 163        run;
 
 NOTE: PROC LOGISTIC is modeling the probability that OBESITY=1.
 NOTE: Convergence criterion (GCONV=1E-8) satisfied.
 NOTE: There were 403 observations read from the data set WORK.TEMP1.
 NOTE: PROCEDURE LOGISTIC used (Total process time):
       real time           0.31 seconds
       cpu time            0.26 seconds
       
 
 NOTE: PROCEDURE LOGISTIC used (Total process time):
       real time           0.00 seconds
       cpu time            0.00 seconds
       
 NOTE: The SAS System stopped processing this step because of errors.
 164        proc logistic data=temp1;
 165        model obesity(event='1')=Waist_Hip_ratio Louisa SBP AGE chol stab_glub glyhb HDL MALE;
 166        contrast 'lrt' SBP 1; AGE 1; chol 1; stab_glub 1; glyhb 1; HDL 1; MALE 1;
                                  ___
                                  180
 ERROR 180-322: Statement is not valid or it is used out of proper order.
 167         run;
 168        
 169        OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 181        

[PaigeMiller]

Thank you. I specifically asked for "We need to see the entire log from PROC LOGISTIC" and not the rest of the program.

Your error is in the CONTRAST statement, which seems to me to be irrelevant to a likelihood ratio test, and so you should remove the CONTRAST statement and run the code again.

[Nouf2]

Sorry about that. I thought it will give you  a better understanding. 

I removed the contrast and it worked. How can I do forward and backward method with maximum model? 

Thank you so much for your help. 

[PaigeMiller]

I think you are asking the wrong person here, as I despise all forms of stepwise regression, and so I never tell people how to do such a thing. Please go to your favorite internet search engine and type in "problems with stepwise regression".

[Ksharp · Accepted solution]

Using option SELECTION= .

model obesity(event='1')=Waist_Hip_ratio Louisa SBP AGE chol stab_glub glyhb HDL MALE   /selection=backward;