Organisation Linked Data Type not working
I have a feature table published on Power BI Service that was previously accessible in Excel via the linked data type. However, for the past month, this connection has stopped working, and I'm unable to fetch data from the table. It appears the issue is related to authentication. Even the promoted dataset cannot be found or accessed anymore. This functionality was working fine earlier. I suspect this is due to a change in the authentication or security layer, because when I try to connect to a Power BI dataset in Excel using the "From Organization" option for a PivotTable, I am prompted to authenticate with Azure Active Directory every time. This suggests that persistent authentication is no longer being maintained. I believe this same issue is affecting the linked data type feature in Excel. Could you please clarify Has there been a recent change in the authentication model? Is there a workaround or fix to restore access to the linked data types from Power BI in Excel? Is the linked data type feature being deprecated? II tested on both current channel and beta channel If it's still working on your tenant, please do let me know or if its broken on your side as well , please update so that I can progress the issue accordingly. Thanks
Power Query is Missing Some Values on Import, Causing IDs to Map to the Wrong Records
Hi. I have a Power Query / Excel issue that I just cannot find a solution to, or a similar problem online. I have a third party cloud based database which I am querying into Excel via an OData connection. The source database contains some duplicate names. As this is real company data, I've created a little demo (screenshot below) to explain the problem. So say I had two companies called Bob's Buns in my original data set. Power Query does not pull the second instance of the name "Bob's Buns", which causes all the names below to line up with the wrong IDs. So now "12" is linked to "Clare's Cakes". You'd think I could just make sure that no two companies in my database have exactly the same name and this would fix the issue, but it doesn't. When I pull the data again after fixing the duplicates, it STILL results in this misalignment error. In fact if I renamed Bob's Buns (ID 11) to "Bob's Buns Bristol", and Bob's Buns (12) to "Bob's Buns Birmingham", Power Query still returns exactly the same result you can see under "What Power Query Returns" below. It's definitely the duplication causing the error, as there were a few instances of this within the database, and the same thing happened. I have done a static export to Excel from the source database, and the results line up correctly with their IDs. However, when I used OData & Power Query, I get this infuriating error. I hope I have explained myself OK! Help very much appreciated!
Making Python-in-Excel worth it
I feel like Python-in-Excel has huge promise. Embedding a Seaborn correlation matrix in an analysis sheet with a couple lines of readable code? Keeping multiple dimensions of data in one cell efficiently? Making models that involve recursion and iteration without all the workarounds of LET/LAMBDA? Yes please!! I've messed around with it a lot but my experience has mostly been negative and from reading comments it sounds like a lot of others have had bad experiences too. A local version that was unlimited for free would be ideal of course. But I'm going to assume for a minute that isn't possible for business reasons. Most of my frustrations relate to these three things. So maybe if we could get these it would be worth it to use? A way to track usage and buy more credits so we don't get stuck with #BLOCKED! with no recourse Improve the xl() function so it works with variables inside it instead of just literals. e.g. foo = "B4" ; bar = xl(foo) currently doesn't work but it should. The error messages need to always tell you what line number the error is in I'm curious if other folks have tried it and agree or disagree with this wish list.
Filter overwhelmed? Not showing all data
Hello! I've developed an application for my workplace, logging absences through the year on a backend spreadsheet and presenting the data on a frontend spreadsheet, with a filter available as managers requested the ability to review data based on names, dates, shift times, etc. It is a lot of data, there are 1833 absences logged over the last 2 weeks. The frontend filter continues to work as expected, and there is practically no delay in presenting the data. However, I've noticed that some of the columns are presented as blank. If I open the backend spreadsheet, suddenly the frontend filter shows all columns as populated. I'm not sure how to lessen the load on my filter, as I am assuming there's excess processing afoot. Any ideas worth a try, thank you! Filter code (crunchy, I know. Eek.) The date variables were my only idea to simplify the want to search from, to, and between dates depending on blank cells: " =IFERROR( LET( fromdate,('BACKEND.xlsm]RECORD'!E:E>=G7)*('BACKEND.xlsm]RECORD'!F:F>=G7), todate,('BACKEND.xlsm]RECORD'!E:E<=H7)*('BACKEND.xlsm]RECORD'!F:F<=H7), tweendate,(('BACKEND.xlsm]RECORD'!E:E>=G7)*('BACKEND.xlsm]RECORD'!E:E<=H7))+(('BACKEND.xlsm]RECORD'!F:F>=G7)*('BACKEND.xlsm]RECORD'!F:F<=H7)), blanks,FILTER('BACKEND.xlsm]RECORD'!A:L,('BACKEND.xlsm]RECORD'!B:B<>"")*('BACKEND.xlsm]RECORD'!B:B<>"Co-Ordinator Entering Leave:")*IF(ISBLANK(B3),1,'BACKEND.xlsm]RECORD'!B:B=B3)*IF(ISBLANK(C3),1,'BACKEND.xlsm]RECORD'!C:C=C3)*IF(LEN(E7)>0,'BACKEND.xlsm]RECORD'!D:D=E7,1)*IF(ISBLANK(I3),1,'BACKEND.xlsm]RECORD'!G:G=I3)*IF(ISBLANK(J3),1,'BACKEND.xlsm]RECORD'!J:J=J3)*IF(ISBLANK(K3),1,'BACKEND.xlsm]RECORD'!L:L=K3)*SWITCH(LEN(G7)+LEN(H7),0,1,10,tweendate,5,IF(LEN(G7)=0,todate,fromdate))), IF(blanks=0,"",blanks)),"No Results") " The data / filter results are listed below, of course personal details removed. Columns A, H, I, and K are all presented as blank until the BACKEND spreadsheet is opened (which defeats the purpose of this filter). BACKEND Data RECORDED Co-Ordinator Entering Leave: Employee Name: Date Called: Absence Start Date Absence End Date Shift Time Affected Actioned? Rostered? Reason for Leave Comments Covered by: TRUE John Smith Don Douglas 08/06/2025 09/06/2025 09/06/2025 0900-1730 NO Personal Leave (PL) Sick, recuperating from injury. Note pending. TRUE Jane Smith Deb Douglas 09/06/2025 09/06/2025 09/06/2025 0600-1430 Sick / Injured Did not feel well, Victoria moved to cover. Victoria Veck Jane Smith Doug Douglas 09/06/2025 09/06/2025 09/06/2025 0600-1430 Sick / Injured Victor Veck FILTER Result ENTERED Co-Ordinator Entering Leave: Employee Name: Date Called: Absence Start Date Absence End Date Shift Time Affected Actioned? Rostered? Reason for Leave Comments Covered by: John Smith Don Douglas 08/06/2025 09/06/2025 09/06/2025 0900-1730 Personal Leave (PL) Jane Smith Deb Douglas 09/06/2025 09/06/2025 09/06/2025 0600-1430 Sick / Injured Victoria Veck Jane Smith Doug Douglas 09/06/2025 09/06/2025 09/06/2025 0600-1430 Sick / Injured Victor Veck
