Check how your country [or US state] performs compared to its peers. And see Cluster Analysis in action — using UMAP Projection & HDBSCAN.

[Update 2020-Apr-19: GitHub repository added, see bottom of the story]

I often wondered about how different countries were compared regarding Coronavirus Case Statistics. Too often, only absolute numbers were taken into account. As a consequence, smaller countries (e.g. Switzerland, Belgium) are overlooked. In this story, I mainly analyze cases in relation to the population size. The hard facts as of today (2020-Apr-16, source JHU) are:

Deaths due to COVID-19 per million inhabitants. Top 25 of all countries, US states, and Chinese provinces having at least a population size of 1 million. Countries are in red, states and provinces in grey.

The state New York is leading this sad top list, followed by the European countries Spain, Belgium and Italy. China has only one province in the top 25 list.

The cheap & managed solution for Mac-only users.

Reduce, Reuse, Recycle — these are the three ‘R’ words behind sustainable living. So why buy new hardware, if the old devices do the job, too?

This is a short 101 on how to set up a cheap and easy-to-use Backup Solution for your Macs and optional Media Hard Disks.

All you need is an external USB Hard Drive for the Backup, plus an (e.g. used) Apple Airport Extreme or Time Capsule, which serves as a cheap, recycled NAS. If you haven’t the latter, you would find it on eBay at 50$+ or so.


All together the setup includes the…

Forget Daily Statistics and follow the Right KPI!

Photo by Sophie Dale on Unsplash

We [and the media!] attach too much importance to the Daily Numbers.
For most countries, the case statistics follow a weekly pattern, with the bottom during the weekend.

Replacing the daily figures and its interpretations by a Rolling Average of the Last 7 Days leads to some decisive advantages:

  • Clear trends instead of shaky, confusing patterns.
  • 60% fewer misinterpretations (i.e. overhasty conclusions).

Let us zoom into the Italian data to better understand this effect:

A comparison of different investments shows that some Bitcoin exposure is needed more than ever. Results plus Code included.

We derive the Asset Allocation which:

  • performed best in the last 6 years
  • showed limited losses during the COVID-19 Crisis
  • is unmanaged (only simple rebalancing once per month)
  • invests only in the S&P 500 index, Bitcoin, and Gold


In terms of risk-adjusted return, 6 years ago, you should have allocated your start capital ($100k) into the mentioned assets by the following weights:

  1. S&P 500: $12.500–20.000
  2. Bitcoin: $7.000–9.000
  3. Gold: $15.000–20.000

plus 53.500–66.500$ in risk-less assets — partially to compensate for the higher risk coming from the Bitcoin allocation. …

Quickstart with Docker & docker-compose. Plus example dashboards showing COVID-19 case numbers.

You wanna deploy your data-driven app using Docker & Docker Compose? Then read on, because this article will get you up-and-running in a few minutes. For both the most wide-spread Data Science / Analytics stacks: Python & R.

In my latest Medium stories, I explained how to set up a data-driven web application for the sake of showing case numbers of the Coronavirus. I created the exact same web application with the following two stacks:

In this article, I will show you how to deploy these apps using Docker & Compose. I…

Zoom and rotate your own COVID-19 —in 20 lines of code.

Screenshot of the App

This brief story is about how you can easily visualise a 3D model of the Coronavirus. The resulting miniature web app allows you to interact with the “virus” via the mouse.

Amidst the Coronavirus crisis we get a lot of case number updates, research and forecasts. Because of the lockdown I have a lot of time to my disposal. Last weekend I created two web apps to download and visualise COVID-19 case numbers (see Medium Posts: Python-Dash, R-Shiny). Now I spent some time on something even more playful: 3D visualisation.

The implementation of this app will be done using R…

Create your own Dashboard Web Application — Today!

This article is less about COVID-19, but more about how to create a useful web application using Python & Dash. It is intended as a simple guide to get up & running with the technology, and makes it easy to do it by yourself!

I wrote pretty much the same article about how to perform this task using R & Shiny, and intentionally wanted to compare these two software stacks. Here I explain the Python-Dash way, but I plan on another article doing a dedicated comparison of the solutions.

To get an idea what the final app would look like…

Create and publish a web app to visualize data on the Coronavirus

At the latest since the Coronavirus COVID-19 is a pandemic, half of the world is paying attention to the case statistics. Me, too. As a Data Scientist, I am curious and so I accessed the case statistics data and created some visualization. With this report, I will tell you how you can do the same by yourself. No magic!

Test it:

This walkthrough explains how to create the web application plus the interactive plots. The open-source technology stack based on R & Shiny is very slim. About 120 lines are all you need to code.

The code can also…

Meinhard Ploner

Lead Data Scientist @ GKN-PM. Strong statistics and Machine Learning background. Loves excellent UI & UX. Creative and always open to new ideas.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store