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

Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.

Reply
Luklaman
Regular Visitor

Bugs

Hi, I think I have found two bugs in PBI and am interested to know if people think they are bugs, or have experienced these issues. I have a PBI report combining 20+ Excel files from multiple different ERP systems that combine all open purchase orders for my organisation. Given the different ERP's I needed a quick solution to identify the internal vendors from the external ones, so I simply created a table (in Excel) with all the vendor names that were internal and entered the word "internal" in the second column. I imported the query and linked this into my master table, and where it could not find that name, enter the word "external". I have an internal vendor called United Airlines, which is 15 characters long, and United Airlines with a space afterwards which is 16 characters long. When I download my source data into Excel and use a if a2=a3,1,"") it confirms they are not identical and also I used a len() to confirm the cell length. In PBI I entered the text.length formula and confirmed they were 15 & 16 characters long. However, when I now link this to the master data table, it informs me that United Airlines has a duplicate entry in my Vendor table. Excel confirms they are not duplicates.....they are different lengths.....PBI confirms they are different lengths.....but thinks they are identical when merging with another table. If I delete one of the United Airlines entries the merging of the two tables is successful, but United Airlines appears as an External vendor. Head bang wall emoji. I can re-write the database to combine the vendor number and site code and use that to determine if a vendor is internal/external, but given the size of data, that is going to take a good day to complete. Whilst that is not a lot of time in the grand scheme of things, it is not something I should have to do. The second issue. The master data has 50+ locations from around the world and we have row level security established. However, for one location, whilst it shows in the Desktop version, once published I am unable to see it. We have changed the name from "London - UK" to "London" and republished. All internet programs are closed. Open IE, login, and we can see "London". However, five minutes later "London" disappears from the list. It still exists in the Desktop file, but not on the published version. The setup of "London" in the RLS Excel file is no different to the other 50+ sites, and all other sites can be viewed without an issue. If we remove RLS entirely, "London" reappears. I have access to all 50+ locations. The real interesting point was that it existed for a period of time after republishing, but then disappeared after then minutes. And yes, we reset to default during that 10 minutes and it stayed there......but disappeared later. Both of these feel-like bugs as there are no logical reasons why PBI has reacted in this manner. It has a group of us banging our heads against a brick wall! I cannot provide any data / examples given data security. Any sensible suggestions anyone!! Other than open a bottle of wine.

3 REPLIES 3

Power Query is case sensitive, so American Airlines is not the same as american airlines. DAX is not case sensitive, so if you load those 2 records into Power BI, they will be treated as being identical. 

you can and should

1. Trim spaces from your column values

2. Clean the columns (this removes non printing characters tha you can't see)

3. Standardise the capitalisation

 

all these steps can be completed from the Power Query transform menu. 

see if that helps. 



* Matt is an 8 times Microsoft MVP (Power BI) and author of the Power BI Book Supercharge Power BI.

Thanks Matt.

 

That has resolved one of the vendors, however I still have two vendors that have not been removed.  I have a vendor called "AMERICAN" and it appears as such in the raw data and the query used to flag Internal vendors.  American is in upper case in both tables and is 8 characters long, but despite this "AMERICAN" remains flagged as an internal vendor.

These things can be challenging because you can't "see" what is happening. Try adding a step in each table query that filters just for the problem item before they are joined. This may help you see what is happening. 



* Matt is an 8 times Microsoft MVP (Power BI) and author of the Power BI Book Supercharge Power BI.

Helpful resources

Announcements
Microsoft Fabric Learn Together

Microsoft Fabric Learn Together

Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City

PBI_APRIL_CAROUSEL1

Power BI Monthly Update - April 2024

Check out the April 2024 Power BI update to learn about new features.

April Fabric Community Update

Fabric Community Update - April 2024

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

Top Solution Authors
Top Kudoed Authors