Kannand only when you are running a mainfrem in Unix mode the Unix rules apply. In the OLD MVS z/OS world it is completely different and nothing like Windos Unix exist. FB VB VSAM have well defined records and do not use a character indicator (termstr) for that. Ebcdic/Ascii code pages adding more comlexity to translations as some might think. ... View more

Andreas das stimmt met unerwartete resultate. Die richtigen name für die LAG function (Oracle is using it with SQL) sollte abe queue sein https://en.wikipedia.org/wiki/Queue_(abstract_data_type) .   Wenn es eine falschen name hat ist die falschen benutzen zu erwarten.  ... View more

https://support.sas.com/rnd/app/examples/stat/GENMODZIP/roots.pdf  Is this helpfull number of roots vs BAP? ... View more

I do not know the analytics design you copied. When the intention of that was only working wiht integers than feeding that with real numbers will be problematic from start. what is what you want to do? Does the code you are using having a fit with that? ... View more

The ACE driver is part of MS-office. MS-office doesn't seems to be installed on the server. It could be done ask you MS support staff. The next thing is downloading that driver and installing it on the server. Ask your MS support staff... Be carefull on installing the same bitness of the driver as you EGuide bitness.  Why has it been missed by the guys doing the SAS installation? Requiremetn interaction wiht MS dedicated file formats....  ... View more

The error: Check whether there are Views involved. SAS-views can point to a physical file that is not there.   The warning: Ignore those messages on compressing. It is an advantage to have that active for big datasets. By not fogetting that as generic setting you can get notes on the very tiny datasets. Performance is important wiht thos big ones. Wiht those tiny ones there is no big deal.   ... View more

of course are real numbers not integers, that are real number for (floating). Maybe the complex numbers are better. ... Never mind you want to give a number as integer, rounded up/down? You could do that before/with the symput No idea whether it makes logical sense as technical it will work.    ... View more

Scott, When implicit is wanted then:   1/ your first check should be on the used functions the must be available in the RDBMS http://support.sas.com/documentation/cdl/en/acreldb/68028/HTML/default/viewer.htm#p0f64yzzxbsg8un1uwgstc6fivjd.htm   2/ You could follow (obs=0) the SQL translation often the SQL is tried behind curtains and when failing all data is copied to the sas-machine. The sqltrace option is a must for analyzing what is happening. A nice paper. http://support.sas.com/resources/papers/proceedings13/072-2013.pdf     3/ dates / datatimes quotings sqlconstdatetime and sqlreduceput are eliminating those sas specifics before trying to convert it to Oracle Sql  http://support.sas.com/documentation/cdl/en/sqlproc/68053/HTML/default/viewer.htm#p12ohgh32ffm6un13s7l2d5p9c8y.htm The quotation in SQL is normally single quotes in ANSI SQL for constant values. SAS is using the double-quotes when allowing macro vars it can be nedessary to stick to those single ones. Thena use %nrstr(%')  for that single quote. The Date-constant should be translated to ansi standards. http://docs.oracle.com/cd/B19306_01/server.102/b14200/sql_elements003.htm#BABGIGCJ   The usage of date/time conventions as of SAS habits can cause problems. I have seen several posts indicating the automatic translation (SAS/access) is failing.       4/ forcing dbmust http://support.sas.com/documentation/cdl/en/acreldb/68028/HTML/default/viewer.htm#p005mmm6mnyvdtn10rxbytrcqjgo.htm          ... View more

I hope you are aiming at the personal /home/<userid> directory at unix. It is has the same goal a "my documents" and %appdata% as known under windows. Having this in mind you can imagine it is quite normal to expect data is left there. It is more troublesome Unix-admins are not seeing that functionality as been locked up in their OS approach.    The problem is the mountpoint filesystem /home  is beloning to Unix-system and not tho the users. Dee a mountpoint-filesystem is some agreed limited by size area as storage (like a usb-drive). What could be done is:     - introduce a group mountpoint filesystem /home/sasusers/ and make that administative the users home location for sas-usage    like  /home/sasusers/<userid> - Make a logical link in everybodies home point to the real data/code shared area's - define some environmentvars at Unix (WorkspaceServer) defining those shared area's so you refer to those using the $  (unix) or ! SAS-file/libnames - Adjust the current dir of a appserver not having the config/lev- location but the users home ~ location (end of users mod sh script)  Expect temporary files and more in ~ but have it limited by sizings ... View more

Mein Tipp: nicht versuchen, alle möglichen Fehler zu programmieren, sondern konzentrieren sich auf die gute gewünschte Situation. Das ist einfacher. Als mögliche Fehler können Sie nicht wissen, was sonst Fehler (unbekannt) sein kann.   Hat mann die -termstmt mit eind aschliessung. Denn wenn das nicht functioniert hat oder es gibt ein SAS error-code oder das resultat (dataein) ist nicht da. Deen wisst mann etwas is nicht gut. ... View more

Murali, it is a interesting question although not a coding or technical working instruction. There are several area-s where the perception on what is going on is important: - Stakeholders assuming the technical implementation of tools is the real problem to get solved - Stakeholders wiht high expectations as of hyping words that all their questions ar getting solved now - The tool implmentators being locked in in their technical stuff without understanding analytic goals. - The tool implmentators not being aligend to business/stakeholders data governance pollicies and data access - The analytical guy trying to understand some (not all) of stakeholder questions - The analytical guy trying to get access to meaningfull data needing some 80% of the time for this. (data governance issues) There is a lot of misunderstandings, but how can you measure those?   No measurements data on that than it is difficult to mak a dashbord or some analyse of that.         ... View more