This sounds like an issue that you should track with SAS Technical Support.  If you are seeing an exception message from EG, it might be a known issue or one that there is a hotfix for.  SAS Tech Support can advise.  They will also help you to gather additional details to diagnose. Are you running EG in English while your Windows session is using a non-English regional settings?  That information might also be of interest to Tech Support. Chris ... View more

Hey there Prabhuk, good thinking of you to add the error message. From this part 'The server encountered an error accessing a source data table.' I would suggest that you maybe deleted the source/drill-through table or such table after you finished building your cubes. To clarify: When you build a cube, you are acutally building the structure of a cube and its metadata for dimensions, measures, aggregations, permissions, etc. The only actual data in the cube itself are some or all aggregations (in this case meaning the values in different levels of intersections) in the cube. To calculate and display non pre-calculated aggregations (or intersections) or to perform a drill through to detail data, the cube does rely on its source/drill-through table to calculate or present data as requested by the user. It seems you have run into just such a situation. What you may want to do is check that you have all source tables registered in the metadata. In the cube studio you might want to edit the cube structure again (use that edit wizard) and check where it says 'Cube Designer - Drill-Through'. Is your table still visible and accessible from there? (try to show the data of the table from the wizard) Alternatively you might want to export the cubes code (via rightclick on the cube) and check the section at the beginning of the code: [...] PROC OLAP    CUBE                   = "/Metadata_Path/OLAP_Scheme/Cube_Name"    DATA                   = Source_Lib.Source_Table    DRILLTHROUGH_TABLE     = maybesomeotherlib.some_table    PATH                   = '/Physical_Path'    DESCRIPTION            = 'Description' ; [...] DATA and DRILLTHROUGH_TABLE might be but are not required to be identical. Both libraries and tables should be registered in the metadata and be accessible. Hope this helps, otherwise let us know what your findings or further errors are. Cheers, Michael ... View more

lets look at your example. Lets see we want to remove East from the left column, and put east as total. How do u do that? I want it this way: Furniture Office Total for East Furniture Office Total for West with a column per year on the right side of course. I couldnt right it properly. but I hope u understand how I want it. Message was edited by: nidhi123 Message was edited by: nidhi123 ... View more

First, your do loop is not defined well for SAS. You can just write [pre] do i=1 to 14; "Do stuff here" end; [/pre] Another thing is that macro variables can't be used within the same data step that they are created. Arrays would probably do what it looks like you are trying to do. [pre] data new; set old; array Exp(14); do i=1 to 14; if Post in ('010') then Exp(i)=Price; end; run; [/pre] But then you will just get 14 colums with the same value. Maybe you could try and explain what you want to do. ... View more

ArtC's point is fundamental to learning SAS programming. Data processing like iterative DO loops is central to the purpose and capabilities of DATA steps. It is possible you want subscripting to select fom a list. That functionality is a typical activity used in an iteration. SAS provides that in DATA steps with ARRAY handling. Not with ampersand. To learn the basics of SAS programming read and try the examples in Documentation. welcome to SAS peterC ... View more

Also, see the topic "SAS OLAP Cube Size Specifications" in the SAS OLAP Server User's Guide.  This topic discusses size specifications for dimensions, hierarchies, levels, measures and properties. ... View more

It is a good and efficient practice to use proc datasets to rename variables, since this will just modify the header portion of the dataset. Sometimes, however, the data set options can be quite handy as well. [pre] /* test data */ data one; x=1; y=1; var1=1; output; x=2; y=3; var1=2; output; x=3; y=2; var1=3; output; run; /* separate and rename */ data two(rename=(var1=newvar1)) three(rename=(var1=newvar2)); set one; if x=y then output two; else output three; run; /* check */ proc print data=two; run; proc print data=three; run; /* on lst Obs x y newvar1 1 1 1 1 Obs x y newvar2 1 2 3 2 2 3 2 3 */ [/pre] ... View more