วีดีโอ

Yes, Dump filters needs more discussion. It will be in the rebuild overview. The are incredibly useful. We have ended up doing a dedicated video - th-cam.com/video/A7sHibnkCUA/w-d-xo.html

Thanks for your comment. I hope you are finding the channel useful. It is too easy for people to forget that each layer can have its own development cycle. That adds so much freedom to a Power BI environment and it is one of the most difficult thing to bring across in videos.

@ I’ve be been a follower for close to 3 years

I'd love to see Power BI VS tableau usage by area!! 😊😊

Thanks, we've moved office and expanding he team. Expansion is what has allowed me to allocate time to the channel again. It has been refreshing seeing how many of the reports packs that were set up for demos are still running... Even this one only needed half an hour's work to change the initial query, the original function still worked! Check Geordie Consulting on LinkedIn to find our in person training/demo sessions as well all in Central Newcastle if that helps.

There's a lot to unpack with that request. If you just want a single view, you can bring all the report content into a single workspace and set up a Dashboard covering them. IF you want to build everything into a single Semantic Model, you can either bring the data together using Dataflows as an intermediary or into a single model... You can copy and paste queries between files using Ctrl + C and Ctrl + V, and then set up the relationships. Measures can be imported using Tabular Editor (2.x works fine), you will no doubt need to tweak them as table or column names may have changed. Our partner consultancy would be able to set up a meeting to get you started. Drop them an email is you are interested office (at) geordieconsulting.co.uk

Yes this video really didn't work as we did it. The high density sampling algorithm in Power BI is locked away. You can see when it is happening and from demos Microsoft have provided it looks to be getting better, but the compromise of a dynamic web based tool was in limiting how many data points it could render reliably. High density scatter plots just don't work as they will always end up dropping some elements (remember maps a really scatter plots), in practice you can use heatspots to get around it, but it is a compromise, in some instances you will lose edge cases that could be statistically interesting or you can lose elements of the "hot spot", there really is no way of configuring the sampling.

Detail is the compromise within Power BI. As you push into more than 1000 datapoints it starts to loose elements (you get an exclamation mark on the visual). Typically this is where the drill feature will help. Use your maps to show numbers at a county, state or country level and then zoom when you drill down to that level show the full map.

Thanks, we're looking at the next series at the moment. If we can find a suitably easy way to build them we will, technically you can do it with notepad but that's a painful way with a learning brickwall rather than curve... Stick with the channel, the new series is finally starting to take shape after a pause for client projects.

Really glad it was helpful. Maps are central to so many of our jobs but they are so often deployed without being fully considered. Opportunities tracking reports are interesting but must be thought about in terms of response capability. Hope it all works for you, that sounds like a great data stack.

Yes, almost every client wants a map at some point. But understanding what is going on can really help you choose what to use when.

You've hit the nail on the head with that one. That was exactly what I wanted viewers to get to. Thin reports are too "Volatile" to really wrap a Change Management process around. That said the are also the part of Power BI that organisations new to the platform struggle to comprehend. The majority of clients that I start working with are duplicating their Datasets not even realising that they can use thin reports. Applying your deployment pipeline to these thin reports is just an "Make-work", in the same way connecting a Dev Deployment Pipeline workspace is just going to force you to follow different schedules otherwise you would be deploying against different source systems. Additionally as the final demo shows editing in situ is all too common in the world of Power BI reports as the pressures of delivery bring out bad habits! In Week 5 we'll build a real world example solution, where Data and visualisations are kept separate. so your thin reports exist within another premium workspace (they need to be Premium to access the content of the Deployment Pipeline). Governance of your Deployment Pipeline will then become much more manageable.

Thanks for the feedback. I try to use the playlists to organise things. But I take your point about the graphics. I'll see what I can do. Ross

Glad it was helpful!

Thanks for your feedback.

Glad you think so! I'll be revisiting Shape files in the coming weeks.

Yes, a limitation of power bi is around 1000 data points. This is a trade off for the online nature of the platform. It is more something to be aware of as it'll be an issue in specific situations rather than all the time.

Also QGIS is a great free desktop GIS if anyone is interested

Thanks for the feedback. Shape maps are not my favorite thing to use in Power BI they tend to be over complex so slow. This video was because I had lots of requests about how to start using shape maps. They are useful but as the longest running preview feature it's nice for people to see how they can get started. I was thinking of re-doing it with a new shape file and I'll be sure to add the enable preview feature step.

Wow, there's a question. Technically tenants are supposed to be separate so no, but I've not experimented enough with the B2B functionality of Azure. From what's there it should be possible to set something up. I suspect you may prefer to go down the Dataflow route for that part however and then pull the result into your datamart. 🤔 So directly no, but something can be done using Azure B2B.

Daisy-chaining is the process of using a dataflow to feed another dataflow. First off you HAVE to have Premium (Capacity or Per User) to be able to use it, so the Workspace it is in must be flagged as Premium. The main challenge is always getting them to run appropriately, my advice is don't go down this route without setting up Power Automate Scripts to manage them. One of the major benefits of daisy-chaining is the rubbish habit I have of using "Base" tables, if you first Dataflows do the RAW data pull and then the data is curated by the second set of dataflows. Source System --> Extract Dataflow --> Curation Dataflow --> Dataset --> Report (Visualisation) Power Automate has a Dataflow Completion trigger, using a condition on the Output of Status (I'll do a video on this) = Success" then allows you to either trigger the first flow again - for failure - or your curation dataflow - for success.

Not with the shape maps. I have heard there may be a plans for it in the Azure Maps visual, but the development plans for Azure Maps are vague.

Yes, getting people to switch to a proper date table early saves so much heartache!

Yes, it's been a busy few months for me and the channel has suffered. I'm preparing a series from "Zero to Hero", going from basics to advanced. I was looking at charticulator for doing some videos on, but I find these solutions tend to be really focused so a solution for one chart just won't work for someone else. I'll keep thinking about it though and see if I can come up with something.

Thanks, I really hate doing Covid reporting, but there is too much nonsense out there. if we can get something meaningful and simple out there then I will have done something good.

Glad it's been of use. The only thing that's still missing is a graduated legend, but you can't have everything can you 🤔

Shape maps are from the shape visual it may still be in preview (it's been there for the last 3 years) There are some supplied with it, but yes you do often have to get creative. New York actually has different shape files for their water towers and Taxis though and the UK is just as bad. You also have to get your data quality nice and high to assign things to the right shape. I'll dig some UK links out to download the shape data and then to make sure it'll work with Power BI

@@GeordieIntelligence That is extremely kind, cheers good sir. Aye, ive found just trying to make a simple shape map fof the 5 towns/citys of Tees Valley is a nightmare, i have the district codes and postcodes, but i cant find any other way other than QGIS - and their interface is just not intuitive. And there's next to no content online on how to do a UK map for PowerBI. 😔

​@@UlyssesHaq Postcodes are a great start. Typically, there is a delay before new postcodes get added to any documentation, so be aware of that for new build areas. Github has some great repositories. These can be an awesome starting point. Postcodes may be the best place to start and then work on getting a shapefile build and the standardisations needed to make it work consistently. Using Postcodes and their Latitude and Longitude, though, can give you some powerful heatmaps. Postcodes - github.com/Gibbs/UK-Postcodes/blob/master/postcodes.csv GeoJSONs (UK) - martinjc.github.io/UK-GeoJSON/ You then need to validate and check your GeoJSON - mapshaper.org/ Video explaining this (sorry about the audio it was a long time ago) - th-cam.com/video/TZXKtl_8yDM/w-d-xo.html

Yes that is the process.science process mining visual, there are a couple available for free on appsource. Week 8 will be a conclusion video, but 9 - 15 will be exploring KPIs and measurements, process mining will be included in that as it is a real game changer. Glad you enjoy the videos, be sure to tell your colleagues about them!