This is episode 117 recorded on September 5th, 2019 where John & Jason talk to Chris Webb, Principal Program Manager at Microsoft on the Power BI CAT Team, about Power Query, data consulting, community efforts, and his new shiny new job at Microsoft.
In the current relationship view window we can delete a data table
from the model.
It would be nice to right click on the data table and click edit
query for the underlying table. This would immediately open up the query editor
for data table changes.
Today you have to switch views to the table view, and click edit table. or navigate to the edit queries button and click edit queries. Doing this opens the queries, but you still have to navigate to the table you want to change.
There
should be a way embedded into the Power BI Desktop or to the Power BI Service
to figure out the actual uncompressed memory footprint.
This
doesn’t seem to be rocket science, and would save a lot of headache. It would
allow analysts to discover how inefficient their data models are. It would
ultimately help the Power BI infrastructure as well by helping people to avoid
loading the Power BI servers with unnecessary workload.
This is episode 116 recorded on September 5th, 2019 where John & Jason talk to Marco Russo about Analysis in general, community efforts, and his new book, The Definitive Guide to DAX, 2nd edition.
As of now, the reports that I create on desktop utilize a lot of time intelligence functions and relative date filters. This is completely fine so long as I view the report on the desktop. Once I publish the report up to the service and then utilize those visuals in dashboards I run into this issue. All of my time intelligence functions (TODAY & NOW) and relative date filters convert to UTC.
This issue causes my data to roll over to the next day
5 hours early which affects all of my subscriptions.
It would be absolutely fantastic if Power BI service could recognize local time zone or even if we could manually set our time zone so that these measure work correctly.
I would like to have the ability to do a “compare tables” or “compare queries” option.
I understand that there are
visuals/queries I can build to compare the DATA within two separate queries but
I want to compare the two queries. I want to know how many columns are in
TableA versus how many are in TableB and which column names are different from
each other. When I have two tables with 150 columns (give or take) and they are
allegedly a copy of each other I want to be able to confirm that all 150 in
columns in TableA match all 150 columns in TableB and that I don’t have 10
unique columns in each query that just happen to come up to 150.
It would be nice to get a row count
difference at the same time, but I do know that I can do the count rows to get
that number of each.
This is episode 115 recorded on August 29th, 2019 where John & Jason talk about the latest News, Ignite update & the August Power BI Desktop news including Grouping visuals, Measure support for the Key Influencers Visual, & some new Data Connectivity.
As Microsoft increasingly supports tailored experiences, we need the ability to operate as a user with equivalent or fewer credentials, including just not being the person who added information.
For example, a site owner who made updates to a site is seeing Updated Pages (with explicit pages underneath) in the left navigation, but has no rights to remove the navigation and didn’t realize that other people weren’t seeing those pages.
Another case was where the SCA had no issue opening a document with embedded documents, but it would not open for site owners and contributors.
In our on-prem environment, I was able to use a test account and assign it with certain rights to verify that those users were going to get the experience I intended, and I was able to do that with a test account. In the SharePoint Online world, for an out-of-the-box solution I’d need a second licensed account for testing in production.
Checking permissions to a site, list, library or item helps, but just isn’t enough for testing the experience on a page in or in a solution.
A potential way to implement this would be to allow me to click on the name I’m logged in as, and in addition to About Me, Sign Out and Personalize this Page, have an option to impersonate a specific user (who has fewer or equivalent rights to me) or a certain level of rights (equivalent to or less than the rights I have).
Some developers in teams help test each other’s solutions, changing who has what rights, but for those of us who largely work independently, we don’t have that luxury.
In the query editor, queries can be grouped (and sub grouped). However, back to Power BI desktop, under the Fields pane, the queries are listed without grouping. Need to retain the same grouping. Also, when publish the dataset, need to be able to hide some query groups from the user.
This is episode 114 recorded on August 21st, 2019 where John & Jason talk to Chris Finlan, Principal Program Manager for Power BI, about his experience at SPFest, whats new and exciting in Power BI, and some specifics around Paginated Reports in Power BI.
This is episode 113 recorded on August 14st, 2019 where John & Jason recap all the news that fit to hear from Summer 2019 like Sending refresh notifications to others, Paginated Reports August 2019 Feature Update, Power BI Premium new capacity settings allow for more control over datasets, and more.
We need the user SSO credentials to be passed through Gateway
connections to Denodo Data Virtualisation layer to be able to securely federate
our data/analytics report queries
Currently,
once you connect to a Power BI Service/Online Dataset you cannot add additional
data sources. We have several datasets we’d like to put in the same report, but
can’t because they are over the 1GB limit on size. If we could live-connect to
them all in Power BI Desktop this would help.
This is episode 112 recorded on August 7st, 2019 where John & Jason talk to Christian Wade, Senior Program Manager for Analysis Services at Microsoft, about Power BI in the Enterprise, the updates to Aggregations feature, upcoming major performance enhancements, and more.
The standard slicers have now a visual level filter, but the new
filter experience does not have a similar feature. A correspondent “visual
level filter” allowing to define conditions that make an item visible or
not. in the new filter experience. The reasons are the same that justified
visual level filter in standard slicers.
This is episode 111 recorded on August 1st, 2019 where John & Jason talk about the latest update to Power BI Desktop including Icon sets for table and matrix, Aggregations now being generally available with RLS support, and the Azure Data Lake Storage Gen2 connector.
If you switch to a custom theme, the conditional formatting dialog
still shows the default theme colors. It would be nice to have the custom theme
colors in the conditional formatting color-pickers.
This
is episode 110 recorded on July 25th, 2019 where John & Jason talk to Anna
Chu from Microsoft about all of the opportunities available for community
participation at this year’s Microsoft Ignite Conference.
This is episode 109 recorded on July 18th, 2019 where John & Jason review some community news, and John interviews Reid Havens of Havens consulting about the Power BI community, consulting in the Power BI world, and template apps.
Text
attributes can be ‘searched’ within report slicers. However, integers cannot.
For example, if someone wants to quickly find and slice by a Customer Id, it is
very cumbersome right now – lots of scrolling. A simple search by Customer Id
(as an integer) within the slicer would make it much faster and a better user
experience
This is episode 108 recorded on July 12th, 2019 where John & Jason interview Justyna Lucznik, Program Manager on the Microsoft Power BI team, about the Machine Learning abilities in Power BI, the announcements around AI & ML at MBAS and what is still to come.
Expose a generic function in DAX that takes R/Python script as a parameter to run on data model data. This enables real advanced analytics in Power BI.
1. Add data labels to Key Influencers so we can see
which data points are the outliers.
Currently, the only way to know is to filter
categories out, or to do a scatter plot of the 2 variables under consideration.
Not practical when having many key influencers.
2. Add color and size options for the graphical view
of the influencers, similar to the Scatter Plot visual
-OR-
3. Add the ML capability to show clusters in the data, in addition with the awesome decision tree for the top segments.
Please
extend Machine Learning capabilities to pro users. Small to Medium sized
business may not have the employee count to justify enterprise licenses, but we
can still benefit from general capability advancements in the Power BI
service/product offering.
