Comments on: Diving deep and the different dialects of MDX http://www.panorama.com/blog/2007/05/diving-deep-and-the-different-dialects-of-mdx/ Making businesses better - not just better informed Tue, 09 Aug 2011 17:51:29 -0400 http://wordpress.org/?v=2.8.2 hourly 1 By: Malaya Das http://www.panorama.com/blog/2007/05/diving-deep-and-the-different-dialects-of-mdx/comment-page-1/#comment-4698 Malaya Das Fri, 12 Jun 2009 09:41:31 +0000 http://www.panorama.com/blog/?p=55#comment-4698 Great solution...Even better from Oz.. Great solution…Even better from Oz..

]]> By: Jonathan Stern http://www.panorama.com/blog/2007/05/diving-deep-and-the-different-dialects-of-mdx/comment-page-1/#comment-26 Jonathan Stern Thu, 15 Nov 2007 21:18:26 +0000 http://www.panorama.com/blog/?p=55#comment-26 This is a great example of the rules of MDX. Having worked with both AS and BW mdx sets (not very similar, other than the name) I find these tips extremely useful. Keep the information coming! This is a great example of the rules of MDX. Having worked with both AS and BW mdx sets (not very similar, other than the name) I find these tips extremely useful. Keep the information coming!

]]> By: Oz http://www.panorama.com/blog/2007/05/diving-deep-and-the-different-dialects-of-mdx/comment-page-1/#comment-14 Oz Tue, 12 Jun 2007 09:57:21 +0000 http://www.panorama.com/blog/?p=55#comment-14 Thanks, Mosha, for paying attention to this special case. Indeed whoever wants to see duplicate data for some reason, should add the "ALL" flag (which I didn't mention in order to simplify the rules of the game). Anyway, it's an honor to have a Grand Master like you in our blog! Thanks, Mosha, for paying attention to this special case.
Indeed whoever wants to see duplicate data for some reason, should add the “ALL” flag (which I didn’t mention in order to simplify the rules of the game).
Anyway, it’s an honor to have a Grand Master like you in our blog!

]]> By: Mosha http://www.panorama.com/blog/2007/05/diving-deep-and-the-different-dialects-of-mdx/comment-page-1/#comment-11 Mosha Fri, 08 Jun 2007 05:50:42 +0000 http://www.panorama.com/blog/?p=55#comment-11 This is an interesting riddle given all the constraints. Unfortunatelly, your solution is not a correct one ! It does seem like your trick worked, but you missed one important detail from the definition of the Generate function. Ever since first spec of MDX came out, it always said, that the Generate function by default removes duplicates from the set. Therefore, if either [True Set] or [False Set] had duplicates, your solution will give wrong results. The correct one should add the ALL flag to the Generate function as following: { Generate (Filter ( { [Measures].CurrentMember }, Condition), [True Set], ALL) , Generate (Filter ( { [Measures].CurrentMember }, NOT Condition), [False Set], ALL) } P.S. When you play chess with GrandMaster, you just might lose :) P.P.S. Also your comment around StrToSet solution that "Besides, what if the sets are huge? Conversion might be time & memory consuming." is not accurate. Since you are passing the names of the sets to StrToSet function, it doesn't matter whether the sets themselves are huge or not. SSAS will quickly resolve the sets by name and there won't be any additional time or memory spent. This is an interesting riddle given all the constraints. Unfortunatelly, your solution is not a correct one ! It does seem like your trick worked, but you missed one important detail from the definition of the Generate function. Ever since first spec of MDX came out, it always said, that the Generate function by default removes duplicates from the set. Therefore, if either [True Set] or [False Set] had duplicates, your solution will give wrong results. The correct one should add the ALL flag to the Generate function as following:

{
Generate (Filter ( { [Measures].CurrentMember }, Condition), [True Set], ALL)
,
Generate (Filter ( { [Measures].CurrentMember }, NOT Condition), [False Set], ALL)
}

P.S. When you play chess with GrandMaster, you just might lose :)

P.P.S. Also your comment around StrToSet solution that “Besides, what if the sets are huge? Conversion might be time & memory consuming.” is not accurate. Since you are passing the names of the sets to StrToSet function, it doesn’t matter whether the sets themselves are huge or not. SSAS will quickly resolve the sets by name and there won’t be any additional time or memory spent.

]]>