How can I modify the length of all variables in a dataset without having to recreate them? Can it be done with Proc Datasets?
I have an old and a new dataset with the same variable names but diffrent lengths, when I merge them I get the warning "Multiple lengths specified for the variable xxxx. This may cause truncation of data".
You cannot change the length of and variable in an existing dataset.
That is one of the basics working with tables/RDBMS. Ask this to a DBA of a RDBMS and you will be see big eyes.
As you can manipulate tables very easy with SAS there is an other way. Understand the way the PDV works using the datastep.
What I mean is you are knowing what it should be, use that knowledge.
data <your newname> ;
length <var ..> ; /* set length-s as needed */
set/merge dsn1 dsn2 ; /* will use length as you have set yourself, not of the input datasets */
run;
Proc append is designed for appending it has assumptions in the same way, assuming the base dataset is leading.
You cannot change the length of and variable in an existing dataset.
That is one of the basics working with tables/RDBMS. Ask this to a DBA of a RDBMS and you will be see big eyes.
As you can manipulate tables very easy with SAS there is an other way. Understand the way the PDV works using the datastep.
What I mean is you are knowing what it should be, use that knowledge.
data <your newname> ;
length <var ..> ; /* set length-s as needed */
set/merge dsn1 dsn2 ; /* will use length as you have set yourself, not of the input datasets */
run;
Proc append is designed for appending it has assumptions in the same way, assuming the base dataset is leading.
I specified the length when I merged the datasets and it worked fine! Thank you...
Here is a convenient way if you only need to SET them together.
proc sql;
create table want as
select * from old
outer union corr
select * from new ;
quit;
And If you only want to change the length of variables :
data have1; input name : $2. x $ id; cards; NB sds 1 RN sdft 1 ; run; data have2; input name $ x : $14. id; cards; NB sdsfsddfs 2 RN sdsdfsdfsder 3 BOTH sdsds 4 ; run; /*First,we get the max length of variables in every tables */ proc sql; create table temp as select memname,name,max(length) as max_len from dictionary.columns where libname='WORK' and memname like 'HAVE%' and type='char' group by name order by memname,name; quit; /* then we use sql to change the variable's length fastly and easily. we need to generate the sql code like: proc sql; alter table have1 modify name char(8),x char(14); alter table have2 modify name char(8),x char(14); ......... quit; */ data _null_; set temp end=last; by memname; if _n_ eq 1 then call execute('proc sql;'); if first.memname then call execute('alter table '||memname||' modify '); call execute(name||' char('||put(max_len,8.)||')'); if not last.memname then call execute(','); else call execute(';'); if last then call execute('quit;'); run; /*finally, get what we want */ data want; set have1 have2 ; run;
Xia Keshan
Don't miss out on SAS Innovate - Register now for the FREE Livestream!
Can't make it to Vegas? No problem! Watch our general sessions LIVE or on-demand starting April 17th. Hear from SAS execs, best-selling author Adam Grant, Hot Ones host Sean Evans, top tech journalist Kara Swisher, AI expert Cassie Kozyrkov, and the mind-blowing dance crew iLuminate! Plus, get access to over 20 breakout sessions.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.