Power BI reports can look like classic dashboards, or they can look more like an app. The use case for Power BI is when there is a large quantity of underlying data, but reports can be styled as the developer wants.

This is Australian Government Department of Agriculture, Water and the Environment data: “The reports identify foods that have failed analytical testing or do not meet the compositional requirements of the Australia New Zealand Food Standards Code”. I rearranged each of the 31 Excel reports from June 2019 to December 2021 to match the format of the January 2022 Excel report, but did not attempt further cleaning. The Excel worksheets where data is collected should adopt data validation, limiting what can be entered in cells and assuring consistent formatting. Using Power Query and M to try to catch data entry errors on the way into Power BI would mean checking the Power BI report every month, whereas data validation in Excel would mean that a dataflow could be established, automating Power BI report updates.

I created many tables behind this visualisation so that I could have the data in the order I wanted it. Because there are only a couple of mentions of North Queensland Export Terminal in the data, I combined the category with Abbot Point. I think they are the same place, but I would have asked before I did this if I was working with a team.

I didn’t know the latitudes and longitudes of the Upstream locations so I just added .001 increments to the latitudes. I wasn’t sure if names like “Bargara” and “Bargara (Bundaberg)” represented the same place so I left them distinct. There was some duplicates in the data. Perhaps some places were sampled twice in the one day with the same result, but I have only counted such samples once.

This is the “Queensland Covid Cases by Age Group” data. You can get some way with plain Power BI, though in this case the order of the age categories is wrong. As you can see, an appropriate data model and a bit of DAX code really add to what can be done with Power BI.

Interesting things in the data? Once the numbers started to rise in mid December 2021, it only took a few days for the 20 to 24 age group to overtake the 25 to 29 group. And in the last few weeks the number of children testing positive has first accelerated and now steadied.

Below, I’ve turned the graph around for a mobile sized version, to avoid having a scroll bar. It’s not possible to officially create a mobile version of a report to embed on a website, but within an organisation mobile versions can be shared.

This is the “Queensland Covid Cases” data. These numbers are the positive PCRs but not the RATs. If you search by a particular postcode, the other slicers will show only the relevant LGA and HHS codes, and sources of infection.

The website that this weather data comes from is amazing. I chose these four weather stations as they are near the coast but I don’t know the geography at all; there may be stations which better capture the nature of Storm Eunice.

To see the carriers and destinations of the cancelled flights that were not assigned an ID, start to type Not Recorded in the aircraft ID search box on the Details page.

Slicers are synced between pages, so you can get a visual representation of the numbers of the carriers and the destinations, by going back to the Flights page.