About Kurt_Bremser

Kurt_Bremser · ‎07-25-2014

Test your credentials by logging on to the server via ssh or telnet. If that works, try to run a simple sas program in batch mode and look at the log. Also have your administrator take a look at the spawner log.

Kurt_Bremser · ‎07-24-2014

I wrote that code under the assumption that every seriesID would have its own line. If more than one come in one line, then my method needs to be adapted to that. You could do a first step that slices the line into the parts that represent single seriesID's and the apply my code to the result.

Kurt_Bremser · ‎07-23-2014

I know that this may look more ugly than necessary, but it follows "if in doubt, use brute force": data want (keep=seriesID year period periodName value code text); infile '$HOME/sascommunity/SeriesDataOut.txt' lrecl=4096 truncover; length data_line $ 4096 seriesID $ 20 year 4 period $ 3 periodName $ 20 value 8 code $ 1 text $ 30 ; input; data_line = _infile_; if substr(data_line,1,12) = '[{"seriesID"'; data_line = substr(data_line,15); i = indexc(data_line,'"'); seriesID = substr(data_line,1,i-1); i = index(data_line,'":['); data_line = substr(data_line,i+4); * removes first { from remaining string; i = index(data_line,'},{'); do while (i ne 0); sub_line = substr(data_line,1,i-1); data_line = substr(data_line,i+3); %macro process_subline; i = index(sub_line,'year":"'); year = input(substr(sub_line,i+7,4),4.); i = index(sub_line,'period":"'); period = substr(sub_line,i+9,3); i = index(sub_line,'periodName":"'); sub_line = substr(sub_line,i+13); i = indexc(sub_line,'"'); periodName = substr(sub_line,1,i-1); i = index(sub_line,'value":"'); sub_line = substr(sub_line,i+8); i = indexc(sub_line,'"'); value = input(substr(sub_line,1,i-1),best5.); i = index(sub_line,'footnotes":[{'); sub_line = substr(sub_line,i+13,length(sub_line)-i-14); i = index(sub_line,'code":"'); sub_line = substr(sub_line,i+7); i = indexc(sub_line,'"'); code = substr(sub_line,1,i-1); i = index(sub_line,'text":"'); sub_line = substr(sub_line,i+7); text = substr(sub_line,1,length(sub_line)-1); output; %mend; %process_subline; i = index(data_line,'},{'); end; sub_line = data_line; %process_subline; run;

Kurt_Bremser · ‎07-22-2014

Could it be that you also have the listing destination open with ODS?

Kurt_Bremser · ‎07-22-2014

You might want to introduce a dedicated delimiter into the input data, like the colon. Then you can use the dlm= option in the infile statement and let the input statement do the separation for you. data want; infile cards dlm=':' format pdt_id 3. pdt_name $20. pdt_quantity 2. ; input pdt_id pdt_name pdt_quantity ; datalines; 101:pencil:57 102:parker pen:24 103:hp studion shuffle n:10 104:samsung galexy ace:29 ; run;

Kurt_Bremser · ‎07-22-2014

if "Mydrive" contains blanks or other special characters, you need to enclose the complete path name to your source/log/print files in quotes.

Kurt_Bremser · ‎07-21-2014

The MDX language has no "IN" operator. You will have to use var='x' or var='y' or var='z' logic

Kurt_Bremser · ‎07-21-2014

You do not have write permissions on the directory that is referenced by the libname sasdata. This is most likely because it resides in the home directory tree of the userid used to install SAS. As a "normal" user, you are not supposed to write/delete anything there. If it is not in the installation user home directory tree, you still need write permissions on the directory wherever it resides. PS you won't see any tables that are visible in EG in the MC unless those tables have been imported into the metadata. Libraries that are not assigned with libname metadata are checked for physical presence of .sas7bdat files and may show tables that are not visible in MC.

Kurt_Bremser · ‎07-17-2014

Assuming that SAS is only available on UNIX, you can do that with Enterprise Guide (File - Export). Another way is to do the export with a data _null_; step and file '$HOME/somefilename.txt';. You do all the formatting needed in the put statement(s). Then you transfer with WinSCP, which allows you to use SFTP/SSH; most UNIXen today do not run simple FTP anymore, for security reasons. With .txt, WinSCP will do the necessary character code conversions for you. SASKiwi's method may not work for lack of a FTP server in Windows.

Kurt_Bremser · ‎07-15-2014

Well, then the only thing I can think of is opening a track with SAS TS.

Kurt_Bremser · ‎07-14-2014

Maybe your DB admin can convert the offending fields to suitable strings, which you then parse in SAS.

Kurt_Bremser · ‎07-14-2014

Are the cubes still visible in OLAP Cube Studio? From your .png, I suspect that the cubes have been deleted and only their respective tables still exist.

Kurt_Bremser · ‎07-14-2014

An alternate method to get the data is to let the DB admin unload the data into a structured flat file and read the data from there with a data step with formatted column input. This is what we do around here.

Kurt_Bremser · ‎07-14-2014

If I interpret your .png right, you try to find something in the OLAP Server that also appears in the Workspace Server(s) The Workspace Server (at least in my installation) can only let you access libraries and tables, _not_ cubes. I suspect that you often have cubes with the same name as the table that they were built from, but some tables do not have an "associated" cube. Without a cube, you can't find anything in the OLAP server.

Kurt_Bremser · ‎07-14-2014

Everytime you need information, googling "SAS item" (where item is the word(s) you need information on) will queickly lead you to the relevant SAS Online Help/Documentation pages. It's what I do all the time.

Online Status	Offline
Date Last Visited	Tuesday