Hi , guys!   Recently I am studying the macro code in the autocall macro in SAS. And I found ther is an error in the autocall macro BUILSFMT source code, which is located in directory: !SASROOT\core\macro.   When I try to call this macro:     %buildFmt(formatName=agefmt,lookupTable=sample,startColumn=age,labelColumn=age_label);   And an ERROR occured:     So I check the source code of macro BUILDFMT, and I found the reaon. The author missed two semicolons. Right here:   These two missed semicolons cause the error above. Just like Lora D. Delwiche and Susan J. Slaughter writed in the Little SAS Book: "Even the most experienced SAS programmer will at least occasionally forget the semicolon."   So, next I am wondering the error jsut in my SAS? Or in all of the SAS? So I check my colleague's SAS which is SAS 9.3, mine is SAS 9.4. His also has the error. So, is there this error in your SAS?   ... View more

This function exists in SCL but not in the Base language. We need to to able to do this:   data _null_; if exist('TABLE') then RC=delete('TABLE'); run;       ... View more

  It seems that as SAS was removing access to the SAS launch command by pushing users off SAS desktop and onto EG, they also moved the option to unbuffer the log to the commande line.   In V6 we had command line option -$logflush, then program option UNBUFLOG in V7, then invocation option again in V9 with option LOGPARM.   Not having the option to unbuffer the log output means that when your SAS session fails, it is very likely that the last bits of log information, which are the bits of interest for investigating the failure, will not be available.       ... View more

...should work on datetime fields without having to use the datepart function.  day(datetime1) should work.  Also, date formats should work on datetime fields.  Date9, for example, should show the standard ddmmmyyyy display. ... View more

Please add encryption to ODSpackage. Either add to the filename statement or ods package will do. Now users have to encrypt the zip post-SAS or in a X command. which adds another step to the processing.   Kind regards,   Patrick ... View more

Keyboard shortcuts that moves the caret to the beginning and end of a selection would be incredibly useful in creating keyboard macros, especially for things like quoting or commenting a specific selection. The current shortcut to comment selections is impractical for most daily uses.    Instead of    /*comment blah*/ /*blah*/ /*blah*/   The shortcut could be used to create a macro that does this   /*comment blah blah blah*/     SOLVED! Thanks @CaseySmith! I knew the arrow keys will take you to the beginning or end, but I was trying to express a shortcut that essentially did both simultaneously, for use in the macro recorder. Your solution of just copying the selection and pasting solves this issue perfectly! It seems so obvious now, thank you for pointing that out!   ... View more

Create the dttime format to mirror the various dtdate formats.   So:   data _null_;    dt=datetime();    put dt dtdate.;  * yields current date ;    put dt dttime.;  * yields current time ; run;   It seems obvious to me, but perhaps I'm missing a good reason why this format does not exist? ... View more

Hi Everyone,   I am currently developing some basic routines to find multiple copies  (some 'd say duplicates ) between SAS Datasets, not inside. Here is one possible blueprint for solving this issue.   Problem situation : cloning is not universal   We have tens of thousands of SAS tables (V9) with many candidates for master + n? copies (usually 1 master +  1 copy, sometimes 2 or more copies). As far as I know, the tables were copied using Proc Copy - unfortunately - not any system tool (cp, copy) which would have saved us - the admins - precious time. Alas, Proc Copy even with clone option doesn't provide a bit-perfect copy : try it for yourself, a sas table copied with Proc Copy is not like a file cloned at filesystem level :   => md5 hash keys for two *.sas7bdat files are different even if Proc Compare displays exact identity between the 2 tables.   a SAS dataset is cloned with reference to its content 'sas-wise' : descriptor (attributes values) + data, whereas a file is cloned as regards its sequential bits at block level 'storage-wise'. The page bounds defined at SAS level might not be exactly equal between the two copies, I suspect, but that doesn't matter for SAS Engine (V9). Even in the case of having two data portions are exactly equal, if some table attributes differ in any sense (e g a variable name, label or format/informat etc.) then we should declare the two copies unequal, obviously.   New tool : an optional Table Digital Signature   Considering the lack of tools at my disposal, I am suggesting to provide a new feature for SAS dataset (V9, CAS) :   let's call it, for instance, a Table Digital Signature TDS. A question of definition immediately arises :   - is this new key to assess unique Tables comprising Descriptor + Observations ? -  or Observations only ?   The latter looks more useful to me in practice.   How - and When - to compute the TDS   The TDS would be based upon some hashing algorithm ; sequence of n bits { bit, bit ... , bit } => 1 Hash key like MD5 or SH1, or some other non deterministic tool like a Bloom Filter, for instance, in order to scale up appropriately  ('linearly') the extra computation time needed : a TDS computed at O(n^2) or, at any rate - beyond O(log n), for instance, would be utterly unusable in practice on large datasets.   Usually, a SAS engine is used sequentially from to top to bottom against a SAS dataset (a data set OBS option is thus an exception) and random or direct access (SET instruction with Point) is rather the exception than the rule : why not use this complete pass through the data to compute the TDS with an efficient single-pass (multi-threaded) algorithm ?   Another idea to allocate the TDS computation time adequately might be to add it to the Compress/Uncompress time (if the Data set is compressed, of course) : in a perfect world, this extra stage could cost only some marginal fraction of elapsed time.   Major difficulty : how to reconcile TDS with partial modifications ?   Like the cost of maintaining an index, any partial (in-place) modification of the dataset :   - Proc Datasets + Modifiy - Data + If + Output - Proc Append etc.   would require a recalculation of the TDS. Therefore, adding some non marginal extra time to the modification  if the modified part is small compared to the overall table size : is there some associative hashing function already available to take care of this ? By associative, here (the exact technical term might be different), I mean some strong property like :   Hash(Big Part of table A) + Hash(Small Part of table A) = Hash(full table A)     See https://www.quora.com/Are-there-any-strong-associative-hash-mixing-functions   How to store and retrieve the TDS value   Like the Sort flag, the TDS could be stored ('embedded') in the table Descriptor portion and displayed with Proc Contents and Dictionary View (Vtables) if available.   How to enable/disable the optional feature   Of course, there should exist some System Option to Enable/Disable the TDS calculation at session/library/table level.    Afterthought : clone signatures aren't perfect   A minor trouble with this kind of feature would be the legal status of the TDS calculation : if false positive (or negative) errors are in theory > 0 (eg hash collisions probability > 0) then this would restrict the usage of such TDS to some kind of only partial (non conclusive) evidence. Taking this into account, I am quite confident this would definitely improve our SAS usage no matter what ! 🙂   If you've liked the idea, thanks for your support ... View more

Because the macro language is not typed (i.e. all values are text strings, rather than being defined as numeric or character) both %Eval and %SysEvalF must do some "guessing" to determine whether to treat text string as a character value or a numeric value.  I would propose to add %EvalC which would always to character evaluation (even if the argument looked like a number) and %EvalN which would always do numeric evaluation (and if the argument could not be interpreted as a number, would return an error).   For example,   1. %put >>%eval(5.0>15)<<; Returns true, because %eval does not know about decimals.  When it sees the decimal, it decides the argument is a character string and does a character (alphabetical) comparison.  This is probably not then result intended by the person who wrote the SAS code.     2. %put >>%sysevalf(5.0>15)<<; Returns false.  %SYSEVALF knows that decimals are numbers and therefore does a numeric comparison as intended.  %SYEVALF also knows about SAS date literalals, and lots of other numbers, which is good.     3. %put >>%sysevalf(793221D360=)<<; Fails (on linux 9.3) with an overflow error.  Here, the user intended a character comparison to check for null, which should return false.  But %SYSVALF saw a number written in exponential notation (D is interpretted same as E) which caused the overflow.   I think these examples all result from the flexibility designed into %EVAL and %SYSEVALF to handle both character values and numeric values.  I would prefer an additional pair of functions which did NOT have this flexibity.   So %EvalN(5.0>15) would return true, because it would know that decimals are numeric (and dates and exponential notation and everything else %sysevalf knows).  But %EvalN(b>a) would return an ERROR, because the values could not be intrepreted as numeric.  Similarly %EvalN(1e2=100) would return true, but %EvalC(1e2=100) would return false, because %EvalC would do a character comparison despite the fact that the argument looks like a number.    Without these functions, if I want to force %EVAL to do a character comparison, a hack workaround is to add a text character so that it cannot look like a number, e.g. %EVAL(Q&a=Q&b).  But that's ugly.  And I don't think there's a good way to force %EVAL to do a numeric comparison or error on a character value  (without inserting my own logic to catch character values).   As a developer, if I write a %IF statement (with implied %eval) or explicit %EVAL / %SYSEVALF, in my design mind I know whether I am intending a numeric comparison or a character comparison.  It would be useful to be able to communicate this intent in my code, rather than let %EVAL / %SYSEVALF guess as to my intent.   Would welcome thoughts/comments from others. ... View more

Hello SAS,   I would like to see the names of those datasets.   Or maybe it is already included in SAS and I cannot find it 🙂   Cheers, Eric ... View more

Due to cybersecurity restrictions, our SAS Site Administrators have enforced system options like SOURCE, SOURCE2 and MPRINT on every SAS session to make sure that every single line of submitted code is printed on the SAS Log. Options SOURCE and SOURCE2 pose no particular problem.  But a locked-in MPRINT option has the unwanted effect of multiplying the number of lines of Macro Code being printed on the Log, every time a Macro is invoked. Having a system option that displays the Macro Code only once (regardless of how many times the Macro is called in the program) would be very helpful to reduce the number of lines displayed on the Log. MPRINT is a debugging option that works well as it does. The new option could be named somewhat differently (maybe MLOG ?). ... View more

Parsing strings into variables is a common task, and it takes a lot of coding to loop over a string, pull out the elements and convert them to variables of the right type. It would be nice to have something that works like the Parse Arg instruction in REXX, so a string could be passed with a template similar to the INPUT statement.   So I have a wish for the new year, a new Call Parse Routine. It would not fit into the general SAS syntax as a function, because the purpose is to create several new variables or assign new values to existing variables instead of returning a single value, but it should fit in implemented as a Call Routine:                   call parse (string,template);   The template is a string containing parse instructions with the same syntax as arguments to the Input statement excluding the trailing @ and @@, which wouldn't make sense in this case. A possible expansion of the syntax could be a way to specify an array name, so any array - including a temporary array - could have values assigned to all variables in one call.   ... View more

PROC SQL is a very powerful way to create macro variables but as far as I can only create LOCAL variables unless I %GLOBAL the names before running SQL.  It would be nice to add a GLOBAL option to the INTO statement so that macro variables are created as GLOBAL. ... View more

It would be very convient to be able to specify a %GLOBAL macro variable and an initial value with a single statement similar to how you can establish a READONLY variable but have it updateable. %GLOBAL variable=ABC; rather than haveing to code... %GLOBAL variable; %LET variable=ABC; ... View more

I occasionally need to break a string into several components such as A_B_C_D becoming A_B_C   and D. The last fragment is easily found by Scanning with a -1 but for the rest, I wind up reversing and left justifying the string and then doing a Substr, another reverse, and another left justification. My suggestion is to add a modifier to Scan so that starting at some point in the string (reading right to left), the returned value is the rest of the string. It would look something like   RestofString = Scan (String , -1 , '_' , 'rest' );   ... View more

I would like some built in functionality to support common tasks associated with working with IPv4 and IPv6 addresses.   For examples of common tasks, I'd personally like to see a function to convert IPv4 to a numeric value and back. For IPv6, I'd like to see something similar to the inet6_ntoa() and inet6_aton() functions implemented by many other languages, and possibly also the ability to represent them easily in formats required by commonly used geo-location lookup datasets such as Maxmind, etc..   From what I can gather, our best options currently seem to be outlined here on by Chris Hemedingers SAS Dummy blog and on this SAS Community post. ... View more

This suggested NOTE would be issued when, in a DATA step, a variable is retained which is present in an incoming dataset; since such variables are automatically retained, this points to a mistake on the part of the coder who wants to create a new variable, and who will be surprised that "the RETAIN does not work" because the retained value is always overwritten by the next observation that is read. Something like NOTE: Variable XXX is read from dataset YYY and therefore automatically retained. ... View more

At the moment, this data T; A1='2:30:00 pm'; A2=input(A1,time11.); A3=input(A1,anydttme11.); B1='2:30:0 p.m.'; B2=input(B1,time11.); B3=input(B1,anydttme11.); run; yields A1 A2 A3 B1 B2 B3 2:30:00 pm 52200 52200 2:30:0 p.m. . .   ... View more

A lot of times people who write macros have lots of comments and instructions in the text of the macro. I'd like to propose a way of showing those comments within a SAS session.  This would especially be handy when the original code file isn't available or convenient.   %macro foobar(a=,b=); %help to run foobar, supply a, a date field, and b, a numeric field: %foobar(a=bdate,b=anum).  This will return something wonderful; (body of macro) %mend foobar;   So, in SAS you could run %foobar /help or %foobar(help) (or something like that); and it would display the help message. ... View more