About half the time I use %sysfunc() is for calling function countw() and %scan() in a loop.   It'd be nice if I could just use %countw().       ... View more

The purpose of CATXF would be to concatenate formatted values with a delimiter. CATXF(format, delimiter, item-1 <, ... item-n>) If format is not present use the variables format attribute. A null delimiter would be a null delimiter (not an implicit 1-space)   Example: array items (5) _temporary_ (1:5); result = catxf(z4., '', of items(*)); Result 00010002000300040005 ... View more

I suggest the creation of additional methods for hash objects which provide the capability to first "position" a pointer in the search tree (e.g. hash.pos_key()), even if the key is not found, and then find the next key in the sorting sequence (hash.next_key()). The second method should have an option to search "up" or "down", the default being "up". The first method will always return a zero, while the second method will return a non-zero return code if the beginning (while searching "down") or the end (while searching "up") of the hash has been reached. Multiple entries for a given key should be treated as one.   ... View more

32 characters is rather short these days, especially when dealing with external (DBMS, excel, etc) data. Why not allow longer variable names? and labels? and likewise for data members? The data structure would not be affected, just the metadata. The change from 8 to 32 (40 to 256 for labels) characters took place for version 7 almost 20 years ago; time for a new enhancement.   Of course, longer text strings in SAS variables (200 to 32k en V7) and macro variables (no change in V7 iirc) would be nice too, but the metadata upgrade should be both easier and more immediately useful. ... View more

There is a bug in SAS that aggressively limits the length of variable names and cripples SAS's ability to interface with any well maintained literally name database. ... View more

It would be handy if SAS accepted date literals in ISO format: '2018-04-05'd and datetime literals: '2018-04-04T08:34:25.678'dt  (the "T" may be replaced by a blank) in addition to the default DATE9. format. ... View more

The RENAME statement (and dataset options) can now handle renaming lists of variable if both the SOURCE and TARGET names are simple numerical suffix lists.  So I could write statements like these: rename old1-old5=new1-new5 ; rename old1-old5=new6-new10 ; rename old1-old5=old2-old6; But frequently we have other lists of variable names, including just simple space delimited names like you would use in any SAS statement.  I would request that we enhance the RENAME statement to allow the addition of extra parenthesis to rename lists of variables.  So things like this: rename (a b c d e)=(name age sex height weight); rename old1-old5=(a b c d e); rename a--e=var1-var5 ; rename (name age sex height weight)=var1-var5 ; rename new:=var1-var5;   ... View more

There are some common variations on how to create delimited files used by data software, like Excel and Redshift, that are currently not directly supported by SAS.  Of these there are two main ones that I think should be easy to implement and solve the most common problems.   Support for reading and writing files with embedded end-of-line characters.  This is common for CSV files generated by EXCEL.  Currently SAS can only handle files where the embedded line break characters do not exactly match the end-of-line characters.  So SAS can read a file that uses CR+LF as end of line and has embedded CR or LF characters, but not records that have embedded CR+LF combinations.  But when the field is properly quoted this is a format that Excel (and others) do support.    This one is especially annoying since it is difficult to program around this limitation.  Currently to read files like this users need to pre-process the whole file to remove or replace the embedded line break characters.  And for writing files SAS does not currently automatically add quotes around field values that contain the end-of-line characters.  So users might need to use the ~ modify on PUT statements to force SAS to add quotes and end up with files that have more quotes than are really needed.   Support for the use of an escape character To protect special characters like delimiters, CR, LF, quotes and the escape character itself a lot of databases (and programming languages) use an escape character, typically a backslash.  This method of protecting special characters in text is popular with many languages, for example Unix command shells, and has been adopted by major database platforms like Redshift.  Typically this is done instead of adding quotes around values that contain special characters. Some even try to add quotes around the values but instead of doubling the quotes that are in the data they add this backslash character. SAS should be able to read a delimited file that is using an escape character instead of quoting to protect embedded special characters. And it should be able to write files that use escaping instead of quoting to protect special characters. ... View more

You can currently use the dataset option RENAME= to change the NAME used for a variable. It would increase flexibility if you could also change other attributes of the variables, such as the FORMAT, INFORMAT or LABEL using a similar syntax.  To avoid conflict with the existing LABEL= option for setting the member name and to follow a similar naming convention as the RENAME= option we might call these options REFORMAT=, REINFORMAT= and RELABEL=.  It should be possible to do it one both input and output dataset references.   Example: data want; set sashelp.class (reformat=(age=5.2)); run; Here is problem that made me think of this.  I was able to attach formats to variables being created with PROC TRANSPOSE by adding a FORMAT statement, but it generated WARNINGs in the log. 88 %let fmtlist=x xfmt. z zfmt.; 88 proc transpose data=have out=want(drop=_name_) ; 89 by id formset ; 90 id var_name; 91 idlabel var_label; 92 var var_value; 93 format &fmtlist; WARNING: Variable X not found in data set WORK.HAVE. WARNING: Variable Z not found in data set WORK.HAVE. 94 run; NOTE: There were 12 observations read from the data set WORK.HAVE. NOTE: The data set WORK.WANT has 4 observations and 5 variables. ... View more

There are many fiddle websites that promote exchange of ideas and demonstrate sample code in a working environment.   There's a fiddle for that - https://fiddles.io - lists 40 or so, make a site so SAS can be added to the list.   If free to use sounds unpalatable, gate the site to SAS community accounts. Code the whole thing in SAS products because, well, ya got to "Eat your own dog food" (wiki)  Could be a wrapper app around a shackled but open SAS Studio login. ... View more

It would be great if there was a function similar to N() that counted the number of non-missing character values. Currently SAS has functions for counting missing character (CMISS) and numeric (NMISS) variables, and nonmissing numeric variables (N). So this seems like a logical hole to fill.   This article suggests development is investigating the possiblity but I wanted to emphasize my interest and hopefull others will as well.   http://support.sas.com/kb/46/235.html ... View more

If one uses an invalid format in %SYSFUNC, the error can go undetected: %let date=%sysfunc(today(),xxx.); %put &=date.; Log: 69 %let date=%sysfunc(today(),xxx.); 70 %put &=date.; DATE=22602 The %SYSFUNC function should issue at least a NOTE to the log, like the PUT function does: 69 data _null_; 70 x = put(today(),xxx.); ____ 484 NOTE 484-185: Format XXX was not found or could not be loaded. ... View more

Please allow the syntax   data T1; set A(gennum=174) - A(gennum=187); run; ... View more

When I make a format, about 99.9% of the time I expect it to be exhaustive of the values that will be encountered.  It would be nice to be able to define a format that would throw an error to the log if a value was not found.   For example:   proc format; value sex 1='F' 2='M' other=_ERROR_ /*pseudo code*/ ; run; data have; input sex ; cards; 1 1 3 ; run;   When the format is used, it would throw an error to the log.  So below PROC step and DATA step would both throw errors, ideally listing the invalid value:   proc freq data=have; tables sex; format sex sex.; run; data want; set have; sexC=put(sex,sex.); run;   I know there are alternatives, e.g., data want; set have; sexC=put(sex,sex.); if sexC="_ERROR_" then put "ERROR: invalid value " sex=; run; But it would be nice to have a way to say when a format is created, "if any value is outside the domain of values defined in the format, throw an error [or warning or note]."   ... View more

I have some processes that are best performed using hash objects, often ending with use of the OUTPUT method to write results to a data set.  But some of my datasets are so large that the correspondng hash objects need too much memory to run.   However, many of these tasks can be done in BY groups.  Conceptually thiis would suggest using the hash object for one by-group, outputitng the results to a data set, then clearing the hash object, and doing the same process all over again for the next by group, followed by outputing it to the same data set.   But the OUTPUT method would only overwrite the data set.  So I am asking for an APPEND method, which would work similarly to PROC APPEND.  This would allow object-oriented processing of many large data sets with a much smaller demand on memory resources.  ... View more

There is still no way to stop a program from executing further statements except by killing the session by aborting it.   Many years ago I kludged together a workaround, but I reckon a proper way to stop a program without generating an error or aborting the session is needed. ... View more

All the major SAS string handling functions have a macro equivalent: %index %substr %scan %upcase %length   Especially when using %scan, one often has to use %sysfunc(countw()) to initialize a macro loop.   So I propose to create a macro equivalent of countw, namely %countw, similar in function to %scan with regards to delimiters and modifiers. ... View more

This would give us, the longest, shortest, and list of values by group.     proc sql; select longest(STR) , shortest(STR) , list(',', quote(trim(STR))) , listnodup(',', quote(trim(STR))) from TABLE group by ID; quit; ... View more

The TERMSTR option is used to define the record delimiter (End of record marker) on flat files. This options does not accept hex values, which is very odd as record delimiters normally aren't standard characters. The syntax   TERMSTR='0c'x   should be supported. The example above was chosen because '0c'x is often used by Hadoop files.   As an aside, this whole option could do with a cleanup as in the following example,   data t; file "%sysfunc(pathname(work))/t.txt" termstr='0c'x; put 'A' / 'B' / 'C' ; run; data _null_; infile "%sysfunc(pathname(work))/t.txt" termstr='0c'; input; putlog _infile_; run; 1- The x is suffix ignored in the first data step's file statement and the delimiter is set to string '0c'.   So the file is 9 bytes long and contains:    A0cB0cC0c   2- In the second data step, SAS somehow finds these 2 nonsensical records:     A0cB   C0cB       ... View more