Hello Reddy19, You wrote: "I'm modifying output format,for example instead of datetime to date etc..comaa. to w.d etc". To change from datetime to date you need a function DATEPART, changing format it is not enough. Please give an example of you data and desired output. Sincerely, SPR ... View more

Hello Darrylovia, I am sure the summing line of your code should look like: [pre] zero=zero+(m[i] ne 0); [/pre] This is a problem of this forum software. Please, see http://support.sas.com/forums/thread.jspa?messageID=27609 Sincerely, SPR ... View more

Hello Mars, These are two common approaches to your problem: a) with a datastep (d1 and d2 should be sorted by X before merging): data c; merge d1(in=d1ind) d2(keep=X A B); if d1ind; by X; run; a) with proc SQL: proc SQL; create table c as select a.*, b.A, b.B from d1 as a left join d2 as b on a.X=b.X ;quit; Sincerely, SPR ... View more

Hello Keen@SAS, You can merge weeks and visits into a new character variable having valies like 2_15 (week=2 visits=15) and plot them on x-axis vertically using VALUE=(ANGLE=90) in your AXIS satetment. Sincerely, SPR ... View more

Hello Grvsinghal, This is a possible solution: [pre] data i; input ID Status $; datalines; 1 A 1 A 1 Z 1 Z 1 A 1 Z 2 A 2 A 2 Z 2 Z run; proc transpose data=i out=t prefix=s; var status; by id; run; data r0; set t; array ss {*} s:; do i=1 to DIM(ss)-1; if ss[i]="Z" and ss[i+1]="A" then do j=1 to i; ss[j]="A"; end; end; by id; run; proc transpose data=r0 out=r(drop=_:); var s:; by ID; run; [/pre] Sincerely, SPR ... View more

Hello lnp, I think that the reason is that normal SAS dataset file has the sas7bdat extension which is different from sas7bcat that format has. In general you can rename file extension of your format file before copying it but it is not a good idea if you have a lot of format files... Sincerely, SPR ... View more

Hello Mars, You need something like this: [pre] proc SQL; create table final.abc as select a.*,b.* /* list all variables you need from two datasets */ from abc as a, salesdata as b where a.date='30Dec2010'd and a.customerNo=b.Custno ;quit; [/pre] Sincerely, SPR ... View more

Hello ExcelSAS, You could also use format month MONYY7.; in your proc TRANSPOSE, then you should get Jun2009 instead of 01Jun2009 like in following example: [pre] data i; d="01Jun2009"d; v=1; output; d="01Jul2009"d; v=2; output; format d date9.; run; proc transpose data=i out=t; var v; id d; format d MONYY7.; run; [/pre] Sincerely, SPR ... View more

Hello AB30, This is a solution: [pre] data i; input ID Grade; datalines; 1 3.2 1 3.7 1 . 1 3.5 2 2.7 2 4 run; proc sort data=i; by ID Grade; run; data r; retain keep; set i; if FIRST.ID then keep=1; if FIRST.ID and Grade = . then keep=0; if keep then output; by ID; drop keep; run; [/pre] Sincerely, SPR ... View more

I am completely confused with the information you submit. In your last post you say: "I want those additional observations in medicare_final but ONLY if the beneficiary_ids match and the admission_date in medicare_final is greater than within 365 days of the admission_date in medicare_proc" 1) It means that admission_date(final) > admission_date(proc) + 365. In you example you wrote: [pre] medicare_proc medicare_final medicare_all id admission_date id admission_date id admission_date 1 03MAR2003 1 03FEB2003 1 03MAR2003 [/pre] According to 1) it should be admission date(final)=03Feb2003 > admission_date(proc)=03Mar2003 + 365=03Mar2003 which is an obvious error. Please, explain. SPR ... View more

Hello SweetPeaIndeed, It is better to use proc SQL in this case like this: [pre] proc SQL; create table medicare_proc_all as select a.*, b.* /* list fields you need from both tables */ from medicare_proc_first as a, medicare_final as b where a.beneficiary_id=b.beneficiary_id and b.admission_date-365 LE a.admission_date LT b.admission_date ;quit; [/pre] Since both datasets contain the same variables like beneficiary_id, admission_date and may be some others it is necessary to list all necessary variables in the select clause. Sincerely, SPR ... View more

Both of these examples are erroneous. It is not such syntax like "sex is missing". SAS generates the following error: [pre] 23 IF sex is missing; __ 388 76 ERROR 388-185: Expecting an arithmetic operator. ERROR 76-322: Syntax error, statement will be ignored. [/pre] HTH SPR ... View more

Hello Bncoxuk, There is an error in your IF statement. Correct version is like this: [pre] IF sex='X' or sex ='' THEN sex='M'; [/pre] Sincerely, SPR ... View more

Hello SASG, This is a solution: [pre] data i; input id date ANYDTDTE9. status $ days; format date date7.; datalines; 112 03/02/09 Approved 4 112 03/03/09 Refferal 3 112 03/04/09 shipped 2 112 03/06/09 shipped 3 113 04/01/08 Refferal 2 113 04/02/09 Approved 1 113 04/03/09 Ahipped 1 run; proc sort data=i; by id date; run; data r; retain drop; set i; if FIRST.id then drop=0; if FIRST.id and UPCASE(status) NE "REFFERAL" then drop=1; if drop=0 then output; by id; drop drop; run; [/pre] Sincerely, SPR ... View more

Hello Alwaz, It is not completely clear to me what would you like to achieve. If you write NODUPKEY instead of NODUP then duplicate by groups will be eliminated from ABC and XYZ and merge will not generate any warnings. Is it your intention? Sincerely, SPR ... View more