how parse null values in every rule

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

    • how parse null values in every rule

      Hi,

      I'm proceeding to arithmetic operations between elements ( 2 - n) and would like to apply a rule every time : if one of the elements is zero valued, then the result is zero.

      for instance :

      ['elt1']=[['elt2']] * [['elt3']] is ok when one = 0, but ['elt1']=[['elt2']] + [['elt3']] or ['elt1']=[['elt2']] - [['elt3']] is not ok ... For my datas zero means 'no data', so the result must be 'no data'.

      I want to create the rules with a java API interface and translate the formula (after validation form) into a PALO rule : users don't know the rule syntax but must have possibility to operate with the elements (time series) of the cube.

      Thank you for any idea.

      I must say that the syntax of the rule, functions is not explained clearly in any document. One must try and try till it works... if someone knows where to find this...
    • Normal users are not supposed to manipulate rules. How do you guarantee that all users use the same calculation syntax ? You also would need a separate cube for each user!

      What is your business requirement? I suppose you are misunderstanding the use and meaning of rules.
    • Thanks for reply.

      My business is time series of economical events for conjecture (tendancies, prospective,...). Each month new datas are collected and some calculations are made between them ( in Jedox : with rules of course)

      The general rule when proceeding to arithmetic combinations between several datas (sum average indices ...) is to return 0 for the data of one month (in fact : no data) till all datas in the calculation chain are collected. The sources of information are multiples and it's really important not to publish a partial result ... without knowing it !!! So I want to control that the rule return zéro as soon as one term is null...

      So rules could help, no ? Till they return 0 as soon as one term is 0 ... but it's not what I understood in the documentation ! Donald (Expert from Jedox Fr) told me it changed in the 5.1 but I think is wrong ! ...

      "Rule is calculated only if at least one the marker arguments on the right side is !=0" So [[S1]]+[[S2]] = S1 if only S2 = 0 ! So it's not what I'm looking for.

      For the users I'm programming a simple graphic interface in java to allow elements combinations. So it's a frame in witch they cannot dammage the cube, and at least I can add "[[ ]]" ...

      The post was edited 1 time, last by tom_zé ().

    • ok, I misunderstood you for the java part.
      Rules sure help you for your requirement. I would set up a small model with few months, small calculation chain and simple rule syntax, eventually without markers in the first run. Thus you can verify your results against some Excel calculation or whatever. You should get a result within short time. Maybe the STET() function helps you.
    • You need to try! Be creative!Many things can be done with PALO. Since it is empty in the beginning it offers a lot of flexibility. (Rules, Support cubes, attributes, subsets etc.)
      It's not just customizing an existing content. If you post an example here (e.g. in Excel) forumers will help you.