The theory behind multiply imputed data only takes estimates and their standard errors and gives combined estimates.  It does not lend itself to combining confidence intervals or p-values (adjusted or unadjusted).  My suggestion would be to combine the estimates using MIANALYZE like shown and then save the p-values to a SAS data set.  You could then use on the p-value adjustment methods available in Proc MULTTEST. Note that not all the p-value adjustment methods are available in MULTTEST that are in GLIMMIX. SAS Help Center: PROC MULTTEST Statement   data outmi; do _imputation_=1 to 3; do trt='test','trt1','trt2','trt3'; do rep=1 to 10; trtn+1; y=1.3+.18*trtn+rannor(123); output; end; end; end; run; proc glimmix data=outmi; by _imputation_; class trt; model y=trt; lsmeans trt/diff; ods output diffs=diff; run; data diff2; set diff; comparison=trt||' vs '||left(_trt); run; proc sort data=diff2; by comparison _imputation_; run; proc mianalyze data=diff2; by comparison; modeleffects estimate; stderr stderr; ods output ParameterEstimates=estimate_ds(rename=(probt=raw_p comparison=test)); run; proc multtest inpvalues=estimate_ds holm hoc fdr bon; run; ... View more

And I noticed that your python function "dfbig.isnull()" only checked the NULL value, you also need to check EMPTY value. Because SAS is unlike database, SAS take EMPTY value as NULL value , but database take them as two different thing . ... View more

If you still need it... you can modify the plot template to customize the y-axis range in a CIF plot.    Here is a default plot:   ods select CIFPlot; proc phreg data=Bmt plots(overlay)=cif;   class Disease;   model T*Status(0)=Disease / eventcode(cox)=1;   baseline covariates=Risk; run;     You can add the following LinearOpts= in the YaxisOpts option to specify your own y-axis range:   proc template;   define statgraph Stat.PHReg.Graphics.Cif;     dynamic title1 title2 title3 title4 xviewMin xviewMax group groupIndex       groupName plotCL transparency _byline_ _bytitle_ _byfootnote_;     BeginGraph;       entrytitle TITLE1 TITLE3;       entrytitle TITLE2 TITLE4 / textattrs=GRAPHVALUETEXT;       layout overlay / xaxisopts=(linearopts=(viewmin=XVIEWMIN viewmax=        XVIEWMAX)) yaxisopts=(label="Probability"        linearopts=(viewmin=0 viewmax=1 tickvaluelist=(0 .2 .4 .6 .8 1.0)));        if (PLOTCL)         bandplot LimitLower=LOWERCIF LimitUpper=UPPERCIF x=TIME / group           =GROUP index=GROUPINDEX modelname="CIF" datatransparency=           transparency;        endif;       stepplot y=CIF x=TIME / group=GROUP index=GROUPINDEX name="CIF";       if (EXISTS(GROUP))         discretelegend "CIF" / location=outside title=GROUPNAME;       endif;     endlayout;     if (_BYTITLE_)       entrytitle _BYLINE_ / textattrs=GRAPHVALUETEXT;     else       if (_BYFOOTNOTE_)         entryfootnote halign=left _BYLINE_;      endif;     endif;     EndGraph;   end; run; ods select CIFPlot; proc phreg data=Bmt plots(overlay range=0,500)=cif;   class Disease;   model T*Status(0)=Disease / eventcode(cox)=1;   baseline covariates=Risk; run;   Delete the modified template to revert back to the default plot:   proc template;   delete Stat.PHReg.Graphics.Cif; run;     ... View more

data have; input ID value1 value2 value3 ; cards; 1 22 33 44 2 33 55 22 ; data want; set have; array x{*} value:; call sortn(of x{*}); max=x{dim(x)}; keep id max; run; ... View more

Thanks!!  That worked.    W ... View more

Let's say you have a dataset with one variable (ICDcode) and 150 observations, one for each of the ICDcodes of interest - call it MY_ICDS   And there is also your main dataset HAVE with multiple variables, including ICDcode.   Then you can use a hash object as follows: data want; set have; if _n_=1 then do; declare hash h (dataset:'MY_ICDS'); h.definekey('icdcode'); h.definedone(); end; if h.find()=0 then flag=1; else flag=0; run; ... View more