Yet another reason I prefer to create reports in Excel and simply publish them.
IMHO, there are *so many* features in Excel that are crucial for business data analysis in custom ways that are independent of the source(s).
PowerBI MUST START trying to get equity with many of the Excel Data Manipulation functions.
Checking with Will Thompson and moving status to backlog
This is a critical issue that is keeping me from migrating several business critical reports from Excel to PowerBI. Having maintained this data for over 7 years, we followed the progression of PowerBI from it's early origins in PowerPivot, creating tons of custom measures with both complex relationships and DAX formulae in their Data Models.
I master my PBI data model in Excel, mostly because of automation and customizations I have therein. Then I import that model into PowerBI.
Making any changes in the data model (adding a column, creating a hierarchy, etc) requires entirely rebuilding the PowerBI reports & dashboards.
Yet another reason I continue to use Excel as my primary reporting UI. 20 years of a mature UI, well worked out with years of revisions.
+1 and a bump if this has fallen off radar.
I developed a fairly complex PowerPivot data model that for a long time I published on a SharePoint site to use with a .bism file as a data source. Basically, the PPDM looks just like an SSAS cube (which I also integrate with).
The complexities of my corporate network made this model unworkable long term, as well as disallowing PowerPivot Automated Refreshes (so, yes, I manually update daily still). So I moved onto Power Query.
For my latest project I had to write entirely new SQL queries to summarize the data that is essentially in that existing PowerPivot data model.... It would have made my life so much easier to simply have referenced that other model than to have to rewrite in SQL, do all the validation necessary to have all the re-sliced data tie out to the existing PPivot reports, and be updated with any maintenance updates on the existing model.