Planning Analytics

Planning Analytics

Get AI-infused integrated business planning

 View Only
Expand all | Collapse all

adding a '1' to all the consolidations

  • 1.  adding a '1' to all the consolidations

    Posted Thu July 18, 2024 01:27 PM

    I have an employee dimension hierarchy where there are around 6 levels in the hierarchy and  each employee is rolled-up at different levels of managers..

    There is a requirement where I need to show employee counts under each manager. For this, at the total consolidation I need to add the counts of all the managers themselves (there is a manager flag). How do I accomplish this?

    The source file I receive has a blank space next to every manager other than direct reportees for the reasons I believe is because they want to show the manager record by itself



    ------------------------------
    Venkata Nori
    ------------------------------


  • 2.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 09:09 AM

    Can you be more specific about the source data and if it is held in a cube?

    You could use MDX, Zero Suppression, add a new measure with a C level rule to identify and count the managers or use a TI to loop the dimension and output counts

    Are there 6 levels in the organisation structure? Do you want to count managers of managers?



    ------------------------------
    Edward Stuart
    ------------------------------



  • 3.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 11:01 AM
    Edited by Venkata Nori Fri July 19, 2024 01:53 PM

    I get the file in .csv format. Source data is all about employee headcount and their leadership information..I have dimensions like geo,dept,entity,date,accounts,job category, time.. I have little experience in MDX and hence trying to use rules. So the thing is at every consolidation I need to add a 1 so that a manager at a level get added to the total count..My file has an unbalanced hierarchy or ragged hierarcy..I use dimensionelement insert and dimensionelementcomponentadd when forming hierarchies



    ------------------------------
    Venkata Nori
    ------------------------------



  • 4.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 11:06 AM
    Edited by Venkata Nori Fri July 19, 2024 01:52 PM

    there is a manager flag. so whereever the flag says 'Y', I am trying to add a 1. The way the source file show data is such that there is a blank cell next to every manager name which means the record is the manager or leader himself



    ------------------------------
    Venkata Nori
    ------------------------------



  • 5.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 11:24 AM

    Be sure not to reveal sensitive information and redact names from any screenshots

    There are a lot of ways to get around this requirement but a carefully crafted C level rule could work. Something that could check if the C level contains child elements.

    If you have not already then (re) familiarise yourself with how rule statements are generated:

    https://www.ibm.com/docs/en/cognos-tm1/10.2.2?topic=data-guidelines-writing-tm1-rules-statements#tm1_dev_adcal_Guidelines_WritingTM1_Rules_Statements

    You could check if a numeric element (e.g. headcount/ FTE) at the consolidation level is greater than 1 then populate a new element (for example 'Manager') with 1 via a rule



    ------------------------------
    Edward Stuart
    ------------------------------



  • 6.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 01:59 PM

    Thanks Stuart..Yes i removed the screenshots..thanks for pointing that..

    Thanks for your pointers on rule documentation and a solution approach.

    I am trying to use the manager flag and then if it is a 'Y', then add a 1 to consolidation..I used below formula but for some reason I am getting 2's in all the cells but it did add 1's to C level elements

    #  ConsolidatedCount( 2, 'Head Count Plan New', 'EXCEL', 'ACTUAL',!Global Leader_New, 'TI Alloc Participant', !Geography_New, !Time, !Job Category New, !Department_New, !Entity, !Employee_New, 'MTD')+1,'TI Alloc Participant');



    ------------------------------
    Venkata Nori
    ------------------------------



  • 7.  RE: adding a '1' to all the consolidations

    Posted Fri July 19, 2024 06:19 PM

    Going back to the base requirement:

    There is a requirement where I need to show employee counts under each manager. For this, at the total consolidation I need to add the counts of all the managers themselves (there is a manager flag). How do I accomplish this?

    For flexibility and control I tend to add a 'dummy' child element to the manager consolidation and use the standard hierarchy consolidations to aggregate the total.

    In the TI process check for the "Y" and if it exists then create a new child element under the 'Manager'

    Alternatives can get quite complex  



    ------------------------------
    Edward Stuart
    ------------------------------



  • 8.  RE: adding a '1' to all the consolidations

    Posted Sat July 20, 2024 04:15 AM

    that dummy child (as a blank)  is what I am already getting from source file and this is exactly how I did insert 'manager_self' next to the blank.

    What this is doing is that it always shows an additional 1 count to all the natural consolidations. for eg: if it is 60 at get layer 5, then it would show 61 and if it is 35 at get layer 6, it would show 36 etc..

    Is it possible to show natural consolidations all the way till the top of hierarchy and at the very top of the hierarchy, to be able to add all manager counts to the final consolidation..



    ------------------------------
    Venkata Nori
    ------------------------------



  • 9.  RE: adding a '1' to all the consolidations

    Posted Sat July 20, 2024 09:36 AM

    I came across a strange situation where I am unable to hide the dummy child using an mdx expression....mdx works temporarily but dummy entries keeps showing up when I expand collapse hierarchy..is there anyway I can hide the dummy child next to the managers?



    ------------------------------
    Venkata Nori
    ------------------------------



  • 10.  RE: adding a '1' to all the consolidations

    Posted Sat July 20, 2024 06:24 PM
    Edited by John O'Leary Sat July 20, 2024 08:43 PM

    If the requirement is " to show employee counts under each manager. " then adding a 1 at each manager level will not achieve that outcome. From what you are describing the need would be to add nothing at level 1 manager and then for all other manager levels except the top you would need to add the number of child managers to the count of the lower levels. For the top level you would not add anything noting that this will add to 1 less than the total employees. 

    I think to do all of this would require some involved C level rules including the use of ElCompn (Or hierarchy version function) and then ConsolidateChildren. All of this is really beyond the scope of a forum question. I would suggest finding an IBM business partner who can help you with the design and build and that will cost money.



    ------------------------------
    John O'Leary
    ------------------------------



  • 11.  RE: adding a '1' to all the consolidations

    Posted Sun July 21, 2024 11:34 AM

    Hello All,

    Thank you all for your valuable time and sending responses. Highly appreciate the same.

    Currently, the way I implemented is that, I have force inserted an element (manager_self) at the time of loading using TI process next to each manager record where there is an empty space. I used the manager flag set to 'Y' for this. 

    And then while loading the headcount, I am checking if the manager element is at N level or C level and accordingly loading either at the manager element itself or 'manager_self'.

    The counts are showing just fine and adding upto the final totals. Only downside  is that at every level the count is showing 1 extra because it is counting the manager himself. I am not quite sure if this is right way of showing or if there is a better way of dealing with this requirement. I  am convinced that we need to show the manager counts themselves in a employee hierarchy but not 100% sure how to explain this to end users. Because they might look at the hierarchy and think, Oh, why am i seeing a number extra under my hierarchy.. This looks little different from any other dimension consolidations.

    Regards,
    Prasad Nori
    Senior Data Scientist
    Indian Software Labs (ISL)





  • 12.  RE: adding a '1' to all the consolidations

    Posted Sun July 21, 2024 11:41 AM

    Yes John..I understood the complexity in using rule for this scenario. Adding a 1 at consolidation is not the only solution



    ------------------------------
    Venkata Nori
    ------------------------------



  • 13.  RE: adding a '1' to all the consolidations

    Posted Fri August 02, 2024 08:07 AM

    Back to what I understood the need to be I have a solution with a 2 Dimension  cube. with no feeders

    • 1st dimension 'Employee' which contains the employees with the rollups through the various managers. This can be ragged
    • 2nd dimension a measure dimension containing an element called 'Reportees'

    No data required, just the structure

    One rule will do it:

    ['Reportees']=C:ConsolidateChildren('Employee')+ElementComponentCount('Employee', 'Employee', !Employee);



    ------------------------------
    John O'Leary
    ------------------------------



  • 14.  RE: adding a '1' to all the consolidations

    Posted Sun July 21, 2024 11:10 AM

    To address the requirement of showing employee counts under each manager while including the counts of all managers themselves, you can use the following approach:

    1. Adjust Hierarchy Levels: Ensure your hierarchy levels are correctly set up in your data source. For each manager, you should have a record in the source file, even if they have no direct reports, as you mentioned.

    2. Add Consolidation Rule: Implement a consolidation rule that sums up the employee counts at each managerial level. This can be done by creating a calculated field or a custom aggregation rule in your reporting tool.

    3. Include Manager Counts: Modify your consolidation logic to include counts of the managers themselves. For example, if you use a flag to identify managers, you can incorporate this into your formula or query to ensure manager records are included in the total counts.

    4. Handle Blank Spaces: If the source file has blank spaces for non-direct reports, make sure your aggregation logic handles these correctly. These blank spaces are typically placeholders for the manager record and should be included in the final counts.

    By following these steps, you can effectively include the counts of all managers in your totals. For more insights, you might want to check out related tutorials or forums where similar issues are discussed.

    This company also does this for referring to its clients by using this : 

    youtube views buy



    ------------------------------
    nadir bhai
    ------------------------------



  • 15.  RE: adding a '1' to all the consolidations

    Posted Sun July 21, 2024 11:44 AM

    Thanks Nadir..Yes I tried 2& 3 above using rule but something is throwing off and it is showing #NA..Mayb i need to dig deeper into rules or understand how an area is defined..

    Sorry to mention this but I did not quite understood the url link above. My apolozies



    ------------------------------
    Venkata Nori
    ------------------------------



  • 16.  RE: adding a '1' to all the consolidations

    Posted Mon July 22, 2024 04:16 AM

    Why not create an element for Manager count and an element for Employee count and a consolidated element to show the Total Count

    The data will not(?) change post load from the source so rules are not a requirement and all the data processing can be done on load



    ------------------------------
    Edward Stuart
    ------------------------------



  • 17.  RE: adding a '1' to all the consolidations

    Posted Mon July 22, 2024 06:56 AM

    Hi

    I would be inclined to used Tis. So firstly a Ti to create the hierarchy itself. Then some Tis that do the following:

    • Master Ti  -  source to be the dimension and return all C items.
    • For each C item in the data tab, call another Ti and pass the C element as a parameter
    • In that other Ti, create your data source dynamically using mdx to return all children (C or N) of that C item.
    • Count 1 for each item in that data section, and then in the epilog convert the value to a string and write against string measure



    ------------------------------
    Chris Hewitt
    ------------------------------