As is almost always the case when building Power BI reports, the first version is just the start of the journey, not the final destination. This article is a follow up from last week where I showed you how I built my COVID Power BI report from scratch. This week I am going to take you through the process I went through of starting to make improvements to build a better version 2.0. Improvements in version 2 include:
- Repoint the local data source to a live online data source (so I don’t need to manually update the source data each day)
- Break the link to the local Excel file that is used to set the phases so I can schedule the refresh of the report without OneDrive or a gateway.
- Build a lead/lag adjustment to time phase align the cases with deaths – you don’t want to miss that one.
- Make the report multi country select, including improving the chart labels
- Improving some of the visualisations so they are easier to read
Here is the video
Here is a Copy of the V2 workbook if you would like to take a look.
What Shall I do Next?
Let me know in the comments below how you would like me to continuously improve this report so it is more useful and user friendly.
Hi, Matt,
In the measure [Report Title], if two or more countries are selected in the Filter Pane, how does the measure [Report Title] shows the selected countries?
Thank you.
Dennis
Hi Dennis,
In the V2 report attached in the article, you will notice that the [Report Title] and the [Chart Title] measures are both using the SELECTEDVALUE() function.
Neither of the measures currently offer an Alternate Result, thus if multiple countries are selected the SELECTEDVALUE() function will return a blank in the space of the country name.
In a situation where you want to have more than one valid country selected at a time, you would be best replacing the SELECTEDVALUE() function with CONCATENATEX().
i.e.
Report Title = “COVID Case Fatality Rate in ” & CONCATENATEX(VALUES(Country[location]), Country[location], “, “)
That’s great, thank you so much Jason.
Hello… well done as it adds more context to the data therein. Any chance you can do something on the exposure sites and its data? As an example the sorting/filtering options are limited. There’s no day of the week which would help. I’ve tried to web scrape the table but my tech skills are limited in this aspect. https://www.coronavirus.vic.gov.au/case-alerts-public-exposure-sites
Unfortunately it looks like that page you linked using some sort of Javascript to generate the page. I would not be able to extract data from that webpage. It’s very frustrating, actually. I much prefer to get the source data and let me at it. Getting the data is the hardest part
Hi Sylvain,
It is possible to extract the information visible on the landing state of the page (i.e. the first 30 rows), using the new Web Table Inference feature.
Unfortunately, at this time, I am not aware of a way of accessing all rows of the table.
Nice
ThX
That was a really enjoyable experience: I will learn what you did in Power BI. And thanks for the mention, too.
We all learn from each other, except those that don’t. ?
Fabulous! Have you considered changing granularity to compare States? I am trying to do so for the US.
This is related to the data source. The data I started with is country data. This is something you can do – you just need to find the data.