Skip to main content
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

Be one of the first to start using Fabric Databases. View on-demand sessions with database experts and the Microsoft product team to learn just how easy it is to get started. Watch now

COVID-19 Coronavirus dashboard - Global pandemic

03-17-2020 06:22 AM - last edited 03-30-2020 17:24 PM

BILLUMINATE
Advocate I
70287 Views
BILLUMINATE
Advocate I
Advocate I

COVID-19 Coronavirus dashboard - Global pandemic

‎03-17-2020 06:22 AM

This Coronavirus dashboard has new insights & user interactions.

  • New measure "Exsiting" = Confirmed - Deaths - Recovered
  • Cumulated & daily view
  • Percentage of deaths / exisiting / recovered
  • Deviation absolute / percent to previous day
  • Select date, continent, country & state
  • 2 Views: By KPI tiles & country hierarchy with heatmap


My progress is on YouTube, where you can follow or redo the development steps. (language: German *Sorry*)

  1. Redone
  2. Redesigned
  3. Remastered
  4. Swiss Lockdown
  5. USA new Hotspot

 

The source is based on the Johns Hopkins Univesity dashboard and dataset on github.

 

Have fun with my dashboard. I look forward to your feedback.

 

Best regards

Nico

 

 

Preview file
341 KB
Message 1 of 5
70,287 Views
czaldumbide
Helper I
Helper I

‎03-23-2020 06:51 AM

Great Work!

I was just wondering how were you handling the scheduled refresh? I've noticed that your data only goes up until a few days before the current date. 

 

Thanks!

Message 4 of 5
69,605 Views
BILLUMINATE
Advocate I
Advocate I
In response to czaldumbide

‎03-23-2020 01:21 PM

Hi czaldumbide,

Thanks!

Short answer: I have to transform the data model each day manually.

Long: 

The source is a Web URL API and should refresh automatically. But the datasets are not perfect. I have 2 options.

  1. Daily .csv-> new table each day
  2. Full time series .csv -> new column each day

I used the full time series because of less tables and joins. For a suitable data model I transformed each table (Confirmed, Deaths, Recovered). With a new columen every day I have to change the transformation steps daily (see attachments).

  1. Increase the number of columns
  2. Chance the data type to Int64

Maybe I overlook a better solution. Do you have an idea?

Best regards

Nico

 

*Update 18.04.2020*

Hint from user africat: Remove parameter "Columen=xx" in Data Transformation > Table > Source

 

 

Preview file
93 KB
Preview file
78 KB
Preview file
71 KB
Message 5 of 5
69,554 Views
cabc_xlorla
Advocate IV
Advocate IV

‎03-17-2020 08:08 AM

Hey mate, great work on building this dashboard. I think the colors are suited for the current situation of confirmed case, recovered and very sadly the deaths...

I have a quick question: is the "Drilldown player" a custom visual built from your self?

Again. I liked the distribution and alignment of the visual. Congrats!

Message 2 of 5
70,230 Views
BILLUMINATE
Advocate I
Advocate I
In response to cabc_xlorla

‎03-17-2020 01:12 PM

Hi cabc_xlorla,

 

the "Drilldown Player" is custom visual from

https://blog.pragmaticworks.com/power-bi-custom-visuals-drilldown-player

Great for the outbreak simulation. But I want dashboard without any third party components (only standard). Maybe I will remove it.

 

Thanks for your Feedback. I will improve the dashboard in the near future.

Nico

Message 3 of 5
70,167 Views