Hitachi Vantara Pentaho Community Forums
Results 1 to 2 of 2

Thread: Search only match first element when dimension nameColumn use the same description.

  1. #1
    Join Date
    Apr 2014
    Posts
    18

    Default Search only match first element when dimension nameColumn use the same description.

    I have the following level declaration at a Payers dimension hierarchy:

    <Level name="payer" visible="true" column="payor_id" nameColumn="payor_name" ordinalColumn="payor_id" type="String" uniqueMembers="false" levelType="Regular" hideMemberIf="Never" caption="Payer">
    </Level>


    Here is the problem. In some cases we are transferring data from the Master Data to the Warehouse were we have records with unique column (Key element) , but the same description at the nameColumn, as you can see with the following sample:

    column= 1234 nameColumn=John Smith
    column= 1235 nameColumn=John Smith
    column= 1236 nameColumn=John Smith
    column= 1237 nameColumn=John Smith

    When performing a search, the MDX will look like this:

    select NON EMPTY {[Measures].[total_payments]} ON COLUMNS,
    NonEmptyCrossJoin([d_organization.h_organization].[business_entity].Members, {[d_security.h_security].[1]}) ON ROWS
    from [payments_cube]
    where {[d_payers.h_payer_group_type].[GROUP1].[GROUP_TYPE_1].John Smith],
    [d_payers.h_payer_group_type].[GROUP1].[GROUP_TYPE_1].[John Smith],
    [d_payers.h_payer_group_type].[GROUP1].[GROUP_TYPE_1].[John Smith],
    [d_payers.h_payer_group_type].[GROUP1].[GROUP_TYPE_1].[John Smith]}

    The payer (last level in hierarchy) is belong to the same GROUP1 and GROUP_TYPE but they are different at the payer level, per column id. Mondrian will only match the first record total and not the rest. Is there a way to configure the schema to use the column identifier to properly aggregate those records in the results set?

  2. #2
    Join Date
    Apr 2014
    Posts
    18

    Default

    As a possible solution to this problem, I also try using captionColumn instead of name column. This solve the problem with the filtering, however it brings the key "column" ids on the results set rather than the descriptive text.
    There should be a flag or configuration at the hierarchy level to designate the filter field, as well as the results field. It should be part of the Mondrian logic to utilize the column as a key for filtering and the nameColumn for display purposes or MDX query.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2017 Pentaho Corporation. All Rights Reserved.