Grouping histogram bars

[errepinna]

Hello everyone,

I'm trying to do something simple. I got different values of the same kind referring to the same entities, and I want to group them in an histogram together.

I made an excel file to visualize what I had in mind:

Only in my case, sales A and sales B are two different tables and the value is not a column but rather a measurement I made with DAX. Is that possible?

[amitchandak]

@errepinna , Put both Location and type on Axis. Expand all levels. Sort on these two. And Switch off concatenate labels.

[errepinna]

My issue is that I have no "Type" column. Each type has a dedicated table so I can't put it on axis.

[amitchandak]

@errepinna ,Create a common location table. Have measure From both Tables and use them in clustered visual

https://www.youtube.com/watch?v=Bkf35Roman8

[errepinna]

I've already made a common table for the "location". Yet if I try to put both measures on values, it just piles them up in a single bar.

 

[v-kelly-msft]

Hi  @errepinna ,

Can you provide a sample .pbix file for test?

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[errepinna]

Here: https://mega.nz/file/CZwBGIjZ#7Mf3VXQYX6Xo7F-bokmwvfZDb9TBXfXssfK7XSlL1Nk

Thanks!

[v-kelly-msft]

Hi  @errepinna ,

Modify your measures as below:

%impactAoverTotA = 
//SUM('A type sale'[A value])/CALCULATE(SUM('A type sale'[A value]), REMOVEFILTERS('type'[Type]))
var _sumA=CALCULATE(SUM('A type sale'[A value]),FILTER(ALL('A type sale'),'A type sale'[Location Code]=MAX('location'[Location Code])&&'A type sale'[Type]=MAX('type'[Type])))
var _sumtotal=SUMX(ALL('A type sale'),'A type sale'[A value])
Return
DIVIDE(_sumA,_sumtotal)

%impactBoverTotB = 
//SUM('B type sale'[B Value])/CALCULATE(SUM('B type sale'[B Value]), REMOVEFILTERS('type'[Type]))
var _sumB=CALCULATE(SUM('B type sale'[B Value]),FILTER(ALL('B type sale'),'B type sale'[Location Code]=MAX('location'[Location Code])&&'B type sale'[Type]=MAX('type'[Type])))
var _sumtotal=SUMX(ALL('B type sale'),'B type sale'[B Value])
Return
DIVIDE(_sumB,_sumtotal)

Then turn off "Concatenate labels" under X axis:

And you will see:

For the related .pbix file,pls see attached.

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[errepinna]

Hello, thank you for your help!

However, this isn't what I'm trying to achieve. I'm already using an all-page filter to select a single "sale type" (RIC, DID, etc). 

On top of that, I'm trying to display the two columns not piled up but separated, so that I can visualize their different height.

So the information is, for example: I want to see, for each location, how much (in %) of total sale A was realized on type DID compared to all Sale A. So I put a filter on the view for type DID and calculate the the %impactonA by removing the DID filter from the denumerator.

I do the same with sale B.

I want to see them close together instead of piled up. I don't care about having the different types on the axis because each type gets his view.

I'm sorry if I wasn't clear about it before.

[v-kelly-msft · Accepted solution]

Hi  @errepinna ,

Is below what you need?

Check the updated .pbix file attached.

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[errepinna]

Worked out! Thanks!