• Black Instagram Icon

©2020 by Fish Wong blog.

  • Fish Wong

On Covid-19 (24 Mar)

This week's covid-19 data analysis consists of covid-19 related tweets analytics and covid-19 data analysis.


Inspired by this analysis from the NYTimes, I decided to investigate a bit more on Chinese media tweets as well.

NYTimes points out that China tends to utilise various state-control twitter accounts to only spread the bright side of the covid-19 situation in China and try to spin the focus to something else. Furthermore, these Chinese media twitter accounts also function to save China's face and to blame other countries for the origin of covid-19.


We can first look at the descriptive statistics from the data.


And these are the most used hashtags used by these Chinese media twitter accounts, apart from hashtags such as "covid-19" or "coronavirus".


I first then created a word cloud that shows the keywords from all of the tweets I scrapped.

And compare it with another word cloud that shows the keywords only from tweets that were posted from February onwards. Words like "cooperation", "fight together", "discharge" and hashtag like "everydayhero" are more prominent now.

Since these 2 word clouds might not appear to have much differences, I also track the number of positive tweets from these accounts. It is obvious that since the mid/ end of February, the number of positive tweets from these Chinese media twitter account surge exponentially.


Lastly here is a word cloud that shows the keywords appear on the positive tweets (tweets that have hashtags "backtowork", "glamorchina", "everydayhero").

We can see these tweets tend to focus on Spring, blossoms and how works in China are resumed.


The covid-19 situation this week


This map shows the confirmed cases around the world in the past 3 weeks.

And here is a static version of the map this week.


Same as last week, I zoom in to the US area and Europe area as well.


Confirmed cases in the US in the past 3 weeks.


A static version of the map for this week's data.


Lastly, here is the map for Europe area.

Countries like Germany, Spain and the UK have the most obvious changes compare to last week.


And for reference, this is the map from last week.


Mortality and recovery rate in different areas


Asia


Europe

Last week, I put the UK on the mild Europe area graph and Norway on the serious graph, the status of these 2 countries changed this week. This aligns with the Europe map from above.

And I think it is important to note that the mortality rate for most of the countries I analysed show an increase in its mortality rate.


The confirmed cases in Portugal started to surge since last week as well. Although it might not be obvious from the graph below, the mortality rate in Portugal is higher than the recovery rate.


Other area

As we can see from the map above, the cases in the US is rising dramatically since last week (and some experts say it will get worse this week), the mortality rate there is relatively low.


Number of confirmed cases in different areas

As pointed out by many already, we should be cautious while reading data related to the confirmed cases, since the number will only increase as more and more people are tested every day. Depending on what testing technique that country is using, the testing numbers per day vary in different areas.


Asia


Europe



Other areas


Thank you for reading!


Further reading:

My previous blog post: On covid-19 (14 Mar)


My previous blog post: On covid-19


My previous blog post: On Coronavirus (2019-nCoV)


#R #DataScience #Society #coronavirus #covid-19