• 1 reply
  • Latest Post - ‏2008-10-30T00:54:34Z by SystemAdmin
151 Posts

Pinned topic table view philosophy

‏2008-10-30T00:43:34Z |

[ said:]


I'm displaying a table in ODM that gathers input data informations collected in a tuple collection .

tuple Tranche
    int inf;//nombre inférieur de palettes de la tranche (inclus)
    int sup;//nombre supérieur de palettes de la tranche (inclus)
    int index;
    float discount;//discount de la tranche
{Tranche} Tranches = {  <1+item(Seuils,i),item(Seuils,i+1),i,-Discount0&#91;1+i&#93;> | i in 0..card(Seuils)-2 : FULLCTR==CS};

The "table" Tranches is not available in ODM views. Therefore, to display these aggregation of input data, I have a simple workaround. I just duplicate the collection in the postprocess phase.

//in post process
{Tranche} bilanTranches = Tranches;

Then, I'm able to use bilanTranches in a ODM view !
There must be a philosophy behind this scheme and I'm eager to understand it !
Updated on 2008-10-30T00:54:34Z at 2008-10-30T00:54:34Z by SystemAdmin
  • SystemAdmin
    151 Posts

    Re: table view philosophy


    [Didier Vidal said:]


    In ODM, you can divide data in two categories:
      * The input of a scenario
      * The output of a scenario

    The data 'Tranches' is computed from the input....  So it falls in the category of the 'output' of a scenario.

    Now, when you automatically create an ODM application from an OPL model:
      * Input is the list of input data that are initialized externally (ie with '=...')
      * Output is
          * The values of the decision variables
          * The values of the variables computed in postprocessing.
    I think this can be overriden if you manually modify some XML files... But the workaround you used is the easiest way.