Assuming (as in your example) you have no by-groups then this should be enough for your purposes   data have; infile datalines dlm=","; input indicador valor; datalines; 1,25 2,25 3,50 ; run; data want; set have; if _n_=1 then valor_acumulado=valor; else valor_acumulado=valor_acumulado+valor; retain valor_acumulado; run; ... View more

If the data is grouped by ID then this should work data want; set have; by id notsorted; Retain Lprice; if first.id then do; lprice=price; lprice=price; end; if Price ne 0 then do; Price_2=Price; Lprice =price; end; else Price_2 = Lprice; drop Lprice; run; Note that if you have multiple levels of grouping variables you may be able to use the FIRST. processing (and often Last.) but you'll want to look at the docmenation about order, by statement options (notsorted descending) and pay attention to which values are changing.    ... View more

Thank you very much! Is working now. Thanks! ... View more

Not sure if you really use them (you need SAS/STAT) or even they are what you need: 1. Using Data step http://www.google.com/url?url=http://people.musc.edu/~simpsona/SASclass/2007%2520SAS%2520Lecture%2520Slides/2007%2520SAS%2520Computing%2520Simulations%2520in%2520SAS.ppt&rct=j&frm=1&q=&esrc=s&sa=U&ei=SuyZU4nSPMXJsQT00ILgDQ&ved=0CDgQFjAG&usg=AFQjCNHjEGTxv0fw5ykS18M8Fkxu4M-GjA 2. Using Proc MI /Logistic /Genmod ( need SAS/STAT) http://www.lexjansen.com/pharmasug/2007/sp/sp07.pdf 3.  Proc GLM (SAS/STAT) http://www.sascommunity.org/sugi/SUGI94/Sugi-94-200%20Wu%20McLean.pdf Haikuo ... View more

I  can use SAS code. I need to join to tables, the columns in table 1 changes all the time and I need to get always all the columns from table 1 and the ones from table to that are not in table 1. PROC SQL;    CREATE TABLE WORK.xxx(label="QUERY_FOR_CONTABILIDAD") AS    SELECT t1.xxx1,           t1.xxx2,    t1.xxx3,           t2.xxx11,    t2.xxx4                FROM WORK.QUERY_FOR_CONTABILIDAD_0001 t1            INNER JOIN WORK.COPIA_DE_LISTADO_DE_CUENTAS_CONT t2 ON (t1.Cuenta_contable = t2.CUENTA);     QUIT; The idea es I hope this more clear. Thanks! ... View more

export it as in cvs format and then to excel ... View more

You can use this general approach (either Data Step method or PROC SQL method) with more than 2 tables. What do you mean by, ". . . I need to have all the variables in table 2"? To help you any further, I think I need to see a sample of your input data and a sample of your desired output. ... View more

Well, if certain conditions are in place ... like both tables contain at most one observation per ID ... you could try it like this: data table1_revised;   set table1;   if name='Specific Value' then name=' '; run; data want;   update table2 (in=in2 keep=id name) table1_revised (in=in1);   by id;   if in1; run; ... View more

Maybe array y[2012:2021]; or array y[10] (2012:2021); ... View more

Prueba de esto, por favor DATA wide;   input id sex $ value1 value2 value3; CARDS; 1 f 40 40 41 2 f 45 44 48 3 m 75 76 77 ; RUN; DATA long1 ;   SET wide ;     value = value1 ;   OUTPUT ;   value = value2 ;   OUTPUT ;   value = value3 ;   OUTPUT ;   DROP value1--value3; RUN; ... View more