Consolidation of time dimension, best way to organize time axis... ?

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

  • Consolidation of time dimension, best way to organize time axis... ?

    I want to compute a consolidation during a period between two dates of the year (for example 01.02 .. 23.07) to be able to compare measures between 01.02.2010...23.07.2010 and 01.02.2011..23.07.2011
    Actually, I have three dimensions for the time axis : year = 2008,2009,2010,2011 / month = 1..12 / day = 1..31

    An idea is perhaps to add a dimension with all combinations of "month" and "day of month" (0101,0102,0103, ... 1228, 1229, 1230, 1231) but it is a redundant with month and day dimensions, and it will make the cube much more greater by adding a dimension with 366 elements...

    Is there another way to get my result ?

    What is the best way to organize the time axis ?

    Another question on the fly : what is the meaning / benefit of "Traiter comme une dimension Temps(MDX)" ? (see MDX.png attached...)

    Thanks for your answers...
    • MDX.png

      3.78 kB, 382×65, viewed 383 times
  • Hi, you may better use one single time dimension in this case. Check for example the following approach:…t=time+dimension#post7277

    You will have to modify it since you are calculating values on the day level.

    You may still keep your original cube with three time dimensions for other analysis purposes. Your original cube may then use only Year and Month as time dimensions? Ths the cube with the day level may serve as a detail cube.

    Regarding your second question:

    Some third-party frontend tools (e.g. Bissantz Delta Master) require the specification of a measure and time dimension in the source database (here Palo).