In comparison with my previous post, this one was an absolute breeze to write. Just an idea, some little tweaking with ChatGPT and GitHub CoPilot. Both come from the daily little miseries of daily with Power BI. One is replacing fields in a visual, the other one is finding the visual.json file of certain visual to see how to tackle that bit of configuration you want to change. Let’s get to it.
Continue Reading..
This is quite crazy. I was trying to write a quick blog post to get back in swing over here, so I thought of a stream lineddd approach set maximum and minimum values for axis to properly show data labels, with a calc group and so on. Then it was impossible for me to sshow an example of the problem I was trying to solve. But then worse. I realized that you can now use visual calculations for those! So my approach was dobsolete before I presented it to the world. Of course I wanted to do something, and as usual I’m very much about automating stuff. Would it be possible to automate the creation of visual calculations?? Well, for that I would need to really parse the visual.json files of the new pbir reports. I have played around modifying the report layer before, but this time I had to do it for real. Instead of creating new stuff, I wanted to look at was already created! Time to tackle all those little jsons.
In this post I’ll explain how I authored a script that detects broken references in PBIR reports, asks the user for replacements, and then fixes the report. However, this is just the excuse, the way I wrote it opens the door to pretty much any script you can think of to modify the report layer. Let’s get started!
Continue Reading..
Hello again!
Before life gets in the way, let’s try to wrap this up. If you have not red the first part, please do so before reading this one as some of the outputs of that part are necessary for the process we’ll see here. In this post I want to tackle the actual report. Since it’s not really data what we’ll translate I’ll call it metadata, but it’s not metadata either. We’ll extract all the titles, subtitles, text boxes, as well as which fields are used where, and crucially what’s the display name that is being used for a particular visual. Of course we may find a field for which no specific display name has been set, but we’ll use the original field name as display name and we’ll take it from there. But enough of all this talking, let’s get to it!
Hello again! This article is part of a series, so if you just discover this, I strongly recommend you start from the beginning. Let’s continue digging through the visual.json files. Today we’ll extract the fields used in visuals and we’ll figure out a way to extract the the first column of field parameters, where we set up the names we want to use. Yes, we’ll need to translate these as well.
Hello again,
after some inactivity the blog is flourishing again. It’s not really something I control. When there’s something I want to write in the blog, I suddenly imagine all the bloggers out there about to write the exact same blog post elsewhere, so I just need to do it, and do it fast. This is something I’ve been doing for real, but really applies in general so I think many can benefit from it. I’ve been asked to take *the* report and make it bilingual. But bilingual for real, so this includes everything: Titles, legends, data, everything. This is a topic that has been addressed a number of times, and yes, there are demos out there. However, the use cases presented in general are either very simple or they start with a project that from the start was known that it had to be bilingual. The reality is, though, rather more complex. I’ll try to write as I progress through this project once I have figured each one of the steps. Let’s do it.
Hello again!
I’m back with the other use case I have in mind for Report-layer scripts. If you haven’t read yet my previous blog post I recommend you do this now as we’ll be building on that with a more complex use case. When I first managed to develop C# in visual studio I rushed to write something I could not build by adapting code snippets only. I made a script that iterated the model relationships and built measures to detect invalid relationships. Not only that, if you build the proper table with the column of the «many» side of the relationship with the measure, you can identify who are the offending values and take action. The script is cool as it is, but again, building each of the tables is quite a pain. Let’s see if we can automate this one. This time I’ll be writing as I author the script. Let’s see how it goes.
Hello!
Today’s post is quite special for me as I wished that this day would come. When I wrote the article to generate the measures for a dynamic data problems button there was a section that read «So, let’s build this button then. This part, unfortunately cannot be scripted, so we’ll need to work a little bit.». Well, maybe this is a thing of the past. Let’s see how can we script this part too. Before going diving in though I want to thank Xavi Paterna for helping me out with his tremendous knowledge of C# programming.
Hello there!
Today’s post comes from an actual requirement that I had heard before but never handled directly. There is some nifty DAX and modeling so lets get to it!
UPDATE: I was kindly asked by Kerry Kolosko to make a Work-Out-Wednesday challenge for this chart, and when I did, I realized that it’s much more simple than I laid out on this article. I keep as proof that finding the best approach is an iterative process! Building the model from scratch also helps. Check out the challenge (and explanation video) here!
Hello there! Time is crazy and I haven’t been updating the blog as much as I’d like to. But here I am and I hope you enjoy it.
This 2024 I took the decision to step up my involvement at Power BI User Group Barcelona and started organizing online events from international speakers. For those that present in English I even do some live interpretation that is recorded and remastered with the video stream so that we get the renowned presenters in both English and in a not-perfect Spanish translation (check out the events here and the recordings here). But why am I speaking about that? Well, first of all to brag about it because these sessions have been awesome, but also because the most recent one is the starting point for this blog post.
Hello!
The other day on the PUG Barcelona whatsapp group there was a question regarding the possibility of reversing the selection on a slicer, with a single click. Apparently this is something you can do on Qlik and the fellow member was trying to migrate that to Power BI as the users are used to be able to do it. I thought that you can get somewhat close to that with a calculation group, so why not write a nice short article on it. Let’s do it!