GAM non linearity

noetsi

Fortran must die
#1
PROC LOGISTIC DATA=WORK.SORTTempTableSorted
PLOTS(ONLY)=ALL
;
MODEL Calculation (Event = '1')=DURATION_02 Box/
SELECTION=NONE
LACKFIT
LINK=LOGIT
CLPARM=WALD
ALPHA=0.05

/* Start of custom user code */
expb
/* End of custom user code */
;
OUTPUT OUT=WORK.PREDLogRegPredictions(LABEL="Logistic regression predictions and statistics for WORK.QUERY_FOR_ONE_0000")
PREDPROBS=INDIVIDUAL
RESCHI=reschi_Calculation
RESDEV=resdev_Calculation
DIFCHISQ=difchisq_Calculation
DIFDEV=difdev_Calculation
C=c_Calculation
CBAR=cbar_Calculation
H=h_Calculation
DFBETAS=_ALL_
STDXBETA=stdxbeta_Calculation
XBETA=xbeta_Calculation ;
RUN;
QUIT;
/* -------------------------------------------------------------------
End of task code
------------------------------------------------------------------- */
RUN; QUIT;
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Try this out. I was assuming Box was your Box transformation, so I left it out.

Code:
ods graphics on;
proc gampl data=WORK.SORTTempTableSorted plots seed=12345;
    model calculation (event='1') = spline(DURATION_02)
    /  link=logit
       dist=binary
       CLPARM=WALD
       ALPHA=0.05;
run;
ods graphics off;
Not sure if your CLPARM option may mix with this procedure or not.