none
Question about value (measure) filters on a pivot table from power pivot data model RRS feed

  • Question

  • <article class="message-body js-selectToQuote" style="box-sizing:border-box;margin:10px 0px 0px;">
    I'm looking for advice on how to do what I'm trying to do in a more efficient way. I have a power pivot data model with data mainly from power query sources pulling from CSV files. The main fact table has a little over 1 million rows, with several other dimension tables and I have several calculated measures. The data is related to payroll, and I am doing an analysis looking for individual employees who are over certain thresholds in a couple different measures. Example, percent of a certain type of pay is >=20% AND dollars of same type of pay are over $5,000. Once I have that set identified, I'd like to be able to slice and dice by different dimensions like division, department, full time/part time, job type, etc. So current approach is adding most of these dimensions as row items in pivot table created from my data model, and some as filters, then applying value filters at the employee ID level in the row items. In order to get BOTH criteria, I just created a duplicate of the employee ID column in the data model table, and then added both to row items and applied one value filter to each (since as far as I know there is no way to apply more than 1 value filter like that to the same row item in a pivot table). However, this seems to make the pivot table very slow to update when any change is made. Seems to get even worse when the 2 employee ID fields move farther down in the list of row items in the pivot table. Example, maybe I start with those first in the list, then move Division and Job type to the top to see employee counts by division and job, suddenly any change I make to the pivot becomes incredibly slow stuck and hangs at "reading data". I have had it take an hour or more when I have actually been willing to wait it out.

    Is there something in this scenario I could be doing in a different way that would be more efficient? I could accomplish the same thing much faster just creating several different pivot summaries and then doing some vlookups and other manual spreadsheet manipulation, so I feel like I must be doing something wrong. I want to make this more easily repeatable via refreshing my data model without the manual spreadsheet work.

    Some things I have tried:
    I originally had a lot more measures showing in my pivot, so I have narrowed those down to just the few main ones I am analyzing right now
    I have gone through where possible removing the FILTER() function from my measures as I have read that it can be quite inefficient
    I significantly reduced the size of my fact table to just one year of data where before I had 4
    </article>
    <footer class="message-footer" style="box-sizing:border-box;margin-color:#141414;font-family:'Segoe UI', 'Helvetica Neue', Helvetica, Roboto, Oxygen, Ubuntu, Cantarell, 'Fira Sans', 'Droid Sans', sans-serif;font-size:15px;background-color:#fefefe;">
    </footer>
    Friday, January 17, 2020 9:18 PM