Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
ryan1982
Helper I
Helper I

Inconsistent Display in Stacked Column Chart

Dear all

 

While dragging data from two seperated charts to make a stacked column chart, the total amount is different from the number added up to total after adding a dimension. 

 

The origional total amount: 11032

 

 

without_filter.JPG 

 

The total amount after adding a dimension: 11345

 

NA_filtered.JPGOthers_filtered.JPG

 

The relationship between those two tables are

 

Address ----> Many to One - Both <----- Cus

 

Those two tables are:

 

"Address" - contain the most-up-to-date address record for each customer. So, this table contains more data since whether those customers chose  to upgrade their membership or not, their customer IDs will be  in this chart. 

(I added a new group in Power BI: place USA & CANADA to "NA", and the rest countries are in "Others" bucket.)  

 

Address.JPG 

 

 

"Cus" - only has the customers who chose to upgrad their memberships and the year they did so.

 

Cus_Year.JPG  

 

1 ACCEPTED SOLUTION
Anonymous
Not applicable

Hi Ryan,

 

If Address table can have more then one records against one customer, this is the intended behavior.

 

A customer having one address of USA and one of JAPAN will be count (distinctively) as one when you are not splitting the count by country_code. When you add country_code dimension, same customer will be counted in both 'NA' and 'Others'.

 

Solution? Try selecting "Count" instead of "Count (Distinct)" in Value Field where you dropped Customer_ID. Or you will have to keep only one (latest) record per customer in address table.

View solution in original post

2 REPLIES 2
Anonymous
Not applicable

Hi Ryan,

 

If Address table can have more then one records against one customer, this is the intended behavior.

 

A customer having one address of USA and one of JAPAN will be count (distinctively) as one when you are not splitting the count by country_code. When you add country_code dimension, same customer will be counted in both 'NA' and 'Others'.

 

Solution? Try selecting "Count" instead of "Count (Distinct)" in Value Field where you dropped Customer_ID. Or you will have to keep only one (latest) record per customer in address table.

Thanks a lot. I found some customers had more than one addresses in the records (a.k.a "Address" table) either due to move, human error, or other reasons. After removing those duplicate addresses and making one ID only had one corresponding address, those two charts matched. 

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.