About Flip

Flip · ‎07-26-2011

Given that "put (variable) (~) ,,," quotes the value of the variable in the output file. I cannot find anywhere where this is documented. It seems a special format, but I am unable to find a reference so that I may comment my code such that others will know what I am doing.

Flip · ‎07-22-2011

I have an ODS output to CSV. The recipient of this file is complaining that there are 2 cariage returns at the end of the file. Is there any simple way to remove those past the data lines? Otherwise I may have to change this to a data _null_ to satisfy them. Ods Listing Close; Title; Ods Csv File = "test.txt" ; Proc Print Data=Ab_Out Noobs; Run; Ods Csv Close;

Flip · ‎01-28-2011

Thanks, I knew it had to be a Friday afternoon brain death.

Flip · ‎01-28-2011

Have I lost my mind or is somethin wrong here? data _null_; cur_yr = year(today()); cur_day = day(today()); cur_mnth = month(today()); x= catx(put(cur_mnth, z2.), put(cur_day, z2.), put(cur_yr, z4.)); y = compress(put(cur_mnth, z2.)|| put(cur_day, z2.) || put(cur_yr, z4.)); put 'HERE ' cur_mnth cur_day cur_yr ; put x=; put y=; call symput('rundat8', catx(put(cur_mnth, z2.), put(cur_day, z2.), put(cur_yr, z4.))); run; %put &rundat8 ; Results in HERE 1 28 2011 x=28012011 y=01282011 NOTE: DATA statement used (Total process time): real time 0.05 seconds cpu time 0.00 seconds 33 %put &rundat8 ; 28012011 How is the CATX reversing the order of the values? Same result in 9.1 on Unix and 9.2 on WIN 7

Flip · ‎06-17-2010

This sort of approach should do that. proc sql; select name into :keeplst separated by ' ' from dictionary.columns where libname = 'XXXX' and memname = 'YYYY' and (upcase(name) LIKE '%ED') or index(upcase(name), 'C') = 3); end; data new; set old (keep = &keeplst); ... run;

Flip · ‎06-17-2010

"I don't want to count on the variables being in the same order every time" In this case you should stay away from using the variable number. You might be deleting different variables each time. What peter means is if there is a logical way to choose names such as: Select name ......... WHERE (some name pattern);

Flip · ‎06-17-2010

compress(location || height); or cats(location , height);

Flip · ‎06-16-2010

Since you are creating the valistvars table, why not edit that table removing the variables you want to drop, then create a macro variable containing the remaining names with an into clause then using that macro variable in a keep statement? select name into :keepnam from varlistvars ... data ; set original; keep ( &keepnam);

Flip · ‎06-16-2010

Patrick; I am going to point this out because I have seen this in several macros at my current location, and it drives me nuts. %put name&i= &&&name&i; will work only if you are aware of everything in your current environment. It should be %put name&i= &&name&i; Why; && -- resolves to & &name -- if it has not been assigned (and you are assuming it has not) it resolves to "name" inside a complex macro variable like this. If there was a global macro variable set it resolves to that. &i -- resolves to your counter. So if somewhere in your job stream there is a %let name = fred, your log will state "&fred1 not resolved". If you are making macros available to others, this can cause problems.

Flip · ‎06-14-2010

Several. 1. Create a format then weight1= input(put(disease1, $newfmt.), best.); 2. Create a lookup table with disease, and weight then do 3 set statements with a "KEY". 3. Create a lookup table then with SQL create columns as a.weight as weight1, b.weight as weight2, c.weight as weight3 from olddata o, left join lookup a on o.disease1 = a.disease left join lookup b on o.disease2 = b.disease left join lookup c on o.disease3 = c.disease

Flip · ‎06-11-2010

Well, you can't say we didn't warn you. It may work in your simple data, but the 2 exist clauses are not associated no matter if you use OR or AND. Each exist looks to see if a record is in that table based on the one condition you have provided. proc sql; update as a set dept='NO DEPT' where exists (select id from lookup l where a.id = l.id and a.sal = l.sal and a.cd = l.cd and a.dept = l.dept) ; quit; Message was edited by: Flip

Flip · ‎06-11-2010

Try this. You will want to check for end of file etc to avoid warnings, but this works in the simple example I wrote. Data new; set old; counter+1; chkvar = byvar; if (condition met) then do pv = counter -10 to counter + 10; set old point = pv; if byvar = chkvar then output; end; run;

Flip · ‎06-11-2010

You are mixing 2 forms of the CASE opperator. use case when value_held = . then value_held = price * shr_held else value_held end as value_held, OR case value_held when . then price * shr_held else value_held end as value_held, Or just use COALESCE(value_held, price*shr_held);

Flip · ‎06-10-2010

Run; data _null_; x = 24617; put(x, date9.); x = 18263; put(x, date9.); run;

Flip · ‎06-10-2010

That could be because 24617 is May26, 2027 Jan 01, 2010 is 18263

Online Status	Offline
Date Last Visited	‎09-01-2015 07:11 AM

Where is this documented

Extra line at end of ODS

Re: Strange result in data step

Strange result in data step

Re: Deleting variables by varnum

Re: Deleting variables by varnum

Re: REmoving interior Blanks

Re: Deleting variables by varnum

Re: Proc SQL into macro variables

Re: Creating new columns(Alternatives)

Re: Using SAS to freeze panes and add colors to cell on Excel spreadsh...

Re: Trying to join two data sets with a data step

Where is this documented

Extra line at end of ODS

Re: Strange result in data step

Strange result in data step

Re: Deleting variables by varnum

Re: Deleting variables by varnum

Re: REmoving interior Blanks

Re: Deleting variables by varnum

Re: Proc SQL into macro variables

Re: Creating new columns(Alternatives)

Re: Macro variable and DATA step-Part 2

Re: How to extract specific ranges of obs from a data set

Re: Proc SQL: Case Expression.

Re: Getting a value from a Text Entry Control in SAS AF App for WHERE ...

Re: Getting a value from a Text Entry Control in SAS AF App for WHERE ...