About Tom

Tom

Copy and paste from where? One annoying thing that SAS introduced with the "Enhanced" editor is that it preserves trailing spaces on the ends of lines. The old editor they had in Display Manager was better suited for code editing. PS Don't place physical tab characters inside of program files. The bytes saved are not worth it. And ironically they make aligning the text harder. It you do want to include them then make sure they only appear at the beginning of the lines.

Tom

Assuming it is not a bug in EG then I suspect in those cases there are white space differences between the two versions. Perhaps one version had trailing spaces removed. Perhaps one version used CR+LF and the other just used LF as the end of line marker. I have not used this EG feature, but based on other difference displays I have used check if EG has an option to ignore white space differences.

Tom

No need to "convert" it. Just change its name. SAS/Studio has a RENAME option. For some reason SAS/Studio makes a big deal of the extension used on the filename. Not sure why.

Tom

You cannot use in-line data inside a macro. Once the macro is compiled into the word token the concept of a LINE of data is gone. But you don't need the datalines, you need the dataset. So generate the data some other way. data _xtemp; do endrange =-100 ,-75 ,-50 ,-30 ,-20 ,-10 ,0 ,10 ,20 ,30 ,40 ,50 ,60 ,80 ,100 ,125 ,150 ,200 ,250 ,300 ,350 ,400 ,500 ,600 ; output; end; run; You could also use the TO and BY keywords to save some typing. data _xtemp; do endrange=-100,-75,-50 ,-30 to 60 by 10 ,80,100,125 ,150 to 400 by 50 ,500,600 ; output; end; run;

Tom

@karanfil wrote: Thanks for the response. I would like to have the following data sets after merging. And run a paired T-test. (like between BSIZEP and DBSIZE) Data A has a YEARP observation during 2018-2019, and data B has a Year 2020-2021 observation. I really appreciate the help. DATA A AND DATA B MERGED TICKER YEARP BSIZEP TICKER YEAR DBSIZE ACGN 2019 8 ACGN 2020 7 . . . ACGN 2021 7 . . . ACMR 2021 5 ACRS 2019 8 ACRS 2020 9 ADAP 2018 10 ADAP 2020 9 That result does not make much sense. How can you have two variables with the same name? Why did you eliminate the value of BSIZEP from one of the observations? Did you mean you wanted this: If so then just just merge by TICKER. data want; merge a b ; by ticker; run; Note that both datasets will need to be sorted by TICKER before you can use the MERGE statement. You can use PROC TTEST to run a paired t-test. proc ttest data=want; paired BSIZEP*DBSIZE; run; But note that will use only four of the observations since one of them does not have the BSIZEP value. And t-test is is not the right statistic for that dataset since the same TICKER value can have multiple observations. And those observations all have the same value of BSIZEP.

Tom

Just make a NEW list. data have; old="Los Lunas/Corona/Corona/White Mountain/White Mountain"; run; data want ; set have; * trick to force data step to define NEW as same length as OLD ; new=old; * Prime the pump with the first word ; new=scan(old,1,'/'); * when next word is not in NEW then append it ; do i=2 to countw(old); if not findw(new,scan(old,i,'/'),'/','it') then new=catx('/',new,scan(old,i,'/')) ; end; drop i; run; Results 90 data _null_; 91 set want; 92 put (_all_) (=/); 93 run; old=Los Lunas/Corona/Corona/White Mountain/White Mountain new=Los Lunas/Corona/White Mountain NOTE: There were 1 observations read from the data set WORK.WANT.

Tom

Have the user tell the macro where to put the value. Say you make a macro like this: %macro MACRO_WITH_RETURN_VALUE (mvar=return_value /* Name of macro variable to store result */ ); %if not %symexist(&mvar) %then %global &mvar; %let &mvar=Value returned ; %mend; Now the user can call it like this: %* Create the macro variable ; %LET RETURN_VALUE =; %* Run the macro ; %MACRO_WITH_RETURN_VALUE(mvar=return_value) %* See what the result was ; %put &=return_value;

Tom

Looks like it is probably related to this issue with SAS Statistical Graphics settings. https://support.sas.com/kb/67/210.html Either turn off Stat Graphics or fix the paths used by it.

Tom

You need to explain WHY that code works.

Tom

@dylanleong78 wrote: thanks a lot for this! sorry it took me awhile to respond. i wanted to make sure applying this logic in my full implementation works. Note that there is a lot of simplification in the code I posted. When using UNTIL() there is no need for an extra test to insure the DO loop executes at least once. You can add an UNTIL() (or a WHILE()) clause to a normal iterative DO loop. So there is no need to increment your loop counter on your own. You can use CAT... series of functions to simplify building strings from other strings and even numbers. And a comment on comments. If you place the comments that explain the code BEFORE the code instead of after it then it is much easier to understand the code. Plus it will keep you from making lines that are too long for humans to read. With long lines in addition to scanning down the page you have to keep sweeping the field of focus left and right.

Tom

@dylanleong78 wrote: thanks. i tried putting short_code=strip(upcase(substr(nam_1_s,1,2))); but it did not work. then i placed the strip in in_prev=index(last_tmp_cd, short_code); and that worked! The reason that putting the STRIP() function call in an ASSIGNMENT statement did not help is because the trailing spaces were added back when the value was stored into the fixed length character variable.

Tom

Most likely the issue is using INDEX() when you actually want INDEXW() or better still the FINDW() function because of the TRIME modifier it allows. data have; input nam_1 :$50. iso_a2 :$2. ; cards; Ham AL Harriet AL Harry AL Hello AL Hitachi AL Hog AL Hogwarth AL Holler AL Hummer AL ; data want; set have; by iso_a2 ; length short_code $2 list $100 ; retain list; if first.iso_a2 then call missing(list); do index=2 to length(nam_1) until(not found); short_code=upcase(cats(char(nam_1,1),char(nam_1,index))); found = findw(list,short_code,',','it'); end; list=catx(',',list,short_code); run; What is it you want to do when you get to something where there is no unique second letter to use?

Tom

Read the documentation on the FIND() function, in particular look at the various modifiers that it offers. Your first example does not work because of the lengths of your character variables. Since you did not tell SAS explicitly what length to use for them it followed its internal rules for guessing what length to use. So STRING is defined as $17 since the first value you assigned to it was 17 bytes long. And CHAR_SUBSTRING was defined as $12 since the default width for the BEST format is 12. So that means FIND() was unable to located a 3 followed by 4 zeros and 7 spaces in "40000,80000,30000". Here are some things you could try: Use the TRIM(), STRIP() or even CATS() function to removing leading/trailing spaces. Use the T modifier in the FIND() function call. Note: Since you seem to be searching for words you might want to use the FINDW() function instead.

Tom

So others know when they are at risk can you clarify what is the length that causes trouble? For example is it names longer than 200 bytes? Or some other limit? Is it the length of the fully qualified filename? Or just the name of file itself? That is could you not rename the file but just move it to a higher directory so that the total name is shorter.

Tom

You appear to be capturing the return code from the function calls rc = metadata_getattr(loginUri, "UserID", userIdValue); , but you are not testing the return code to check if the function succeeded before checking the value of the data set variable. if lowcase(userIdValue) = lowcase("&user_id") then do; Perhaps the call is failing and this is the cause of the false hits?

Online Status	Online
Date Last Visited	yesterday