Covid-19 Analysis

Covid-19 Analysis

Date of analysis: 14 December 2020

By Rizami Annuar

Top 20 Worldwide (sorted by Cases)

CountryCasesDeathsMortality rate (%)
South Africa860,96423,2762.70

Malaysia and UAE (sorted by Cases)

CountryCasesDeathsMortality rate (%)

The World

CasesDeathsMortality rate (%)

Why I created the Covid-19 analysis?

  • Most Covid-19 graphs analyses are time-series, i.e. the cumulative total is plotted on the vertical y-axis and date on the horizontal x-axis. I haven’t seen graph analysis comparing daily count (on y-axis) and cumulative total (on x-axis).
  • After watching the Minutephysics YouTube video about How To Tell If We’re Beating COVID-19, I had the idea to link the daily count and cumulative total.
  • From my own analysis, I’m beginning to seeing the daily count and cumulative total provide a more realistic leading indicator to eyeballing the trend – Are we improving? Worsen? “Flattening the curve” will happen over time (the “black line” on graph below). Yet, the “blue line” provides greater clarity for immediate trend signal – are we trending up or trending down. Trending down means the Covid-19 controls are effective, whilst trending up means there is a need for better Covid-19 controls.
  • The blue line fits the Covid-19 data to approximate the trend. As daily count declines, the blue line will drop, which is a positive sign. When there is no new daily Cases or Deaths being reported, the blue line will move closer to the horizontal x-axis to form a small letter “n”. That is the goal we want to achieve, the small letter “n” blue line!
  • Graph with black line: Cumulative total on vertical y-axis and time-series on horizontal x-axis.
  • Graph with blue line: Daily count on vertical y-axis and cumulative total on horizontal x-axis.
  • The blue line smoothing is based on local regression (loess) technique (reference).

Comparative: Malaysia Cases Lockdown Day 273

  • Malaysia lockdown: 18 March 2020
  • China Hubei lockdown: 23 January 2020
  • Italy lockdown: 9 March 2020

Given the lockdown started on separate dates, comparatively the Covid-19 controls in Malaysia are effective and the blue curve is trending in the right direction.

Comparative: Malaysia Deaths Lockdown Day 273

Cases Cumulative – Various Selected Countries. From 31 December 2019 to 14 December 2020

How the rest of the world is doing in the fight against Novel Coronavirus Covid-19:

Deaths Cumulative – Various Selected Countries. From 31 December 2019 to 14 December 2020

Correlation between China, Italy and Malaysia Lockdown Day 273

Correlation is a metric that measures the degree of relationship between two variables. The countries’ high percentage of correlation, suggests that they are closely related and predicting each other.

The estimated cases correlation:

  • Malaysia and China: 36.5%
  • Malaysia and Italy: 98.8%

The estimated death correlation:

  • Malaysia and China: 58.2%
  • Malaysia and Italy: 81.6%


  1. European Centre for Disease Prevention and Control; An agency of the European Union; Download today’s data on the geographic distribution of COVID-19 cases worldwide link
  2. RStudio Team (2018). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA link. Taiyun Wei and Viliam Simko (2017).
  3. R package “corrplot”: Visualization of a Correlation Matrix (Version 0.84). Available from link.
  4. Correlation definitions: (i) link (ii) link (iii) link
  5. minutephysics (2020) How To Tell If We’re Beating COVID-19. Available at: link (Accessed: 29 March 2020).
  6. StatQuest with Josh Starmer (2017) Fiitting a curve to data, aka lowess, aka loess. Available at: link [Accessed 31 March 2020].
  7. (2020). Local Regression. [online] Available at: link [Accessed 31 March 2020].
  8. Wicklin, R. (2016). What Is Loess Regression?. [online] SAS Blogs. Available at: link [Accessed 31 March 2020].
  9. Prabhakaran, S., (2016). Loess Regression And Smoothing With R. [online] Available at: link [Accessed 31 March 2020].