Different results for custom column in service

[Alpesh]

Hi, any help/thoughts, as I;m getting different results between Deskptop and Service. The service doesn;t show correct calculation...

Desktop (correct calculation)

Service (wrong calcultion):

Formula:

_1st Accepted Business Hours =

VAR _Start = 'Chat Table Unique'[Unprocessed Time]

VAR _End = 'Chat Table Unique'[Accepted Time]

VAR _Workhours =

SUMX(CALCULATETABLE(

    'Date - Working Hours per Day',

    DATESBETWEEN('Date - Working Hours per Day'[Date],_Start,_End),

    'Date - Working Hours per Day'[Final Working Day factoring in Holiday] = 1

    ),

    MAX(MIN('Date - Working Hours per Day'[Business End],_End) - MAX('Date - Working Hours per Day'[Business Start],_Start),0) *24

)

Return

IF(_Workhours,_Workhours,0)

[Alpesh · Accepted solution]

After alot of fault finding, it turns out that because I was converting in Power query every datetime to the local call center country it completely threw off the calculation when refreshed. So instead I used a DAX formula (TIME()) instead to convert the UTC times into the local timezones for the calculation... and now it works!!

[TheoC]

@Alpesh make sure to mark your answer as the solution so others can find this post 🙂

All the best!

Theo

[TheoC]

@Alpesh  Ahhh yes, that will definitely influence the outputs. Well done on finding the cause!  This post is a great post to showcase the importance of applying a systematic process for data preparation (i.e. preparation / transformation / manipulation). 

Again, great work in finding the cause! This highlights something very important which is often overlooked:  when using Power Query to manipulate / transform data, we will not be able to easily see the log of these changes directly in Canvas view.  

Great stuff @Alpesh!

[TheoC]

Hi @Alpesh 

The DAX measure looks fine and is quite logical. I don't believe that's the issue.  From my experience, there are several reasons why this could be occuring:

1. Time Zone differences between your computer and the browser.

2. Filters applied / not applied in Power BI Service vs Desktop.

3. Row level security (if applicable) depending on how it's set up.

There are potentially other reasons but hopefully one of the above will help resolve.

Let me know how you go.

Theo 🙂