Hi @billi_billi  You can put  model medical=age Race gender / cl solution; ... View more

Hi @abudyak,   Are your samples paired? ... View more

Hi @raja777pharma    Thank you for the clarification. Here is an attempt to create table3. Let me know if does meet your expectations. Best, /* TABLE 3 */ %macro macro2 (prefix = ); data table3_&prefix.; set table1; length TYPE $ 5; TYPE = upcase("&prefix."); if &prefix.CT = "X" then flag_CT = "CT"; if &prefix.MRI = "X" then flag_MRI = "MRI"; if &prefix.OTH = "X" then flag_OTH = "OTH"; TRT_MULTIPLE = tranwrd(catx("_","TRT",flag_CT,flag_MRI,flag_OTH),"_.",""); if TRT_MULTIPLE = "TRT_CT" then TRT_MULTIPLE = "CT_01"; if TRT_MULTIPLE = "TRT_MRI" then TRT_MULTIPLE = "MRI_02"; if TRT_MULTIPLE = "TRT_OTH" then TRT_MULTIPLE = "OTH_03"; keep SUBNUM VISNAME TESTDAT TYPE TRT_MULTIPLE; run; %mend; %macro2(prefix =TEST); %macro2(prefix =EXTRA); %macro2(prefix =MORE); data table3; set table3_test table3_extra table3_more; run; ... View more

Hi @raja777pharma    Hereafter is a piece of code to create table2. NB: I guess there is a typo in the name of variable "TEST_OTH" -> it should be more logical to have TESTOTH (please rename it so that the macro can work).   I don't understand the logic to create table3 ->could you please explain a bit more? In particular, how should we consider the 'TYPE' variable ? -> it seems that the data you displayed in table 3 correspond to type = "MORE" ?   Best,   data table1; infile datalines dlm="09"x dsd missover; input SUBNUM VISNAME $ TEST_DEC $ TESTDAT:date9. TESTCT $ TESTMRI $ TESTOTH $ EXTRA_DEC $ EXTRADAT:date9. EXTRACT $ EXTRAMRI $ EXTRAOTH $ MORE_DEC $ MOREDAT:date9. MORECT $ MOREMRI $ MOREOTH $; format TESTDAT EXTRADAT MOREDAT date9.; datalines; 001 VSIST1 Yes 16Apr2019 X Yes 16Apr2019 X Yes 16Apr2019 X 002 VSIST1 Yes 06May2019 X Yes 06May2019 X Yes 06May2019 X 002 VSIST2 Yes 13Jun2019 X Yes 13Jun2019 X Yes 13Jun2019 X X 003 VSIST1 Yes 22Apr2019 X Yes 22Apr2019 X Yes 22Apr2019 X X 004 VSIST1 Yes 11Sep2019 X Yes 11Sep2019 X Yes 11Sep2019 X 004 VSIST2 Yes 05Nov2019 X Yes 05Nov2019 X Yes 05Nov2019 X 004 VSIST2 Yes 03Dec2019 X Yes 03Dec2019 X Yes 03Dec2019 X 005 VSIST1 Yes 12Dec2019 X Yes 12Dec2019 X Yes 12Dec2019 X 006 VSIST1 Yes 30Oct2019 X Yes 30Oct2019 X Yes 30Oct2019 X X 007 VSIST1 Yes 23Oct2019 X Yes 23Oct2019 X Yes 23Oct2019 X 008 VSIST1 Yes 02Jan2020 X Yes 02Jan2020 X Yes 02Jan2020 X ; run; /* TABLE 2 */ %macro transp (prefix = ); proc transpose data=table1 out=table1_&prefix (rename=(&prefix.DAT=DATE)); var &prefix.CT &prefix.MRI &prefix.OTH; by SUBNUM VISNAME &prefix.DAT; run; %mend; %transp(prefix=test); %transp(prefix=extra); %transp(prefix=more); data table1_int; set table1_test (in=a) table1_extra (in=b) table1_more (in=c); length TYPE $ 5 ; if a then TYPE = "TEST"; else if b then TYPE = "EXTRA"; else if c then TYPE = "MORE"; length TRT $ 8; if prxmatch('/CT/', _name_) then TRT = "CT_01"; else if prxmatch('/MRI/', _name_) then TRT = "MRI_02"; else if prxmatch('/OTH/', _name_) then TRT = "OTH_03"; if col1 ="X" then output; drop col1 _name_; ; run; data table2; set table1_int; by TYPE SUBNUM VISNAME DATE notsorted; if first.date then count= 0; count + 1; if count > 1 then TRT = "MULTIPLE"; drop count; if last.date then output; run;   ... View more

Hi @Mick_lb    You can use the LARGEST() function that returns the xth largest non-missing value.   best, data have; input Note_A Note_B Note_C; datalines; 10 12 14 16 14 8 15 16 12 ; run; data want; set have; Max_note_1 = largest(1, Note_A, Note_B, Note_C); Max_note_2 = largest(2, Note_A, Note_B, Note_C); run; ... View more