COVID data story

Hi all :slight_smile:

Are there some people who are currently working with some COVID data stuff ? or who have an interest for ?

Have you got an opinion to share about those data ?
Where do you find your data ?
What do you find interesting with ?
What do you do as analyses on it ?

Iā€™m really interested by all the methodological questions about collecting data and how to treat them and by how they are presented in different kind of medias as newspapers, speaking news on tv or radioā€¦Also, Iā€™m curious about how counts are made in each countryā€¦and how countries are compared with each other.

Have a nice day <3

2 Likes

Iā€™m somewhat interested on building a toy project related to COVID. Maybe something real-time so I can practise building stuff with Phoenix.

Though, I have no clue where to start to find a constantly changing and open data source. Data sourcing and analytics are definitely not my forte.

Thx for your replying :slight_smile:

Could you precise your idea of toy ? Do you think something like an app which is able to give you some ā€˜real-timeā€™ info about, for example, the spread of infection in your country then in your living area, the number of deathā€¦or something like that ?

By constantly changing do you mean which is regularly updated ?
I saw on linkedin posts that people are working on tracking app and that they are files on github, maybe you could dig this track and try to find with which data bases they work.

Could you explain me what is the main idea under using phoenix framework ? I mean why this one and not another one ? Iā€™m really new into javascript and js libraries/frameworks so consider you are speaking to a martian :smiley:

What is (are?) the part you find the most difficult ?

I have sleepless night with questionning sourcing hahaā€¦
Little story about :

Some years ago, I worked as an encoder student for a big survey (multiple countries) and I had to work on a part of metadata study which consists to analyse a selection of newspapers content from individual country (we are different coders from every country which participated and coded in our first language). The adventure is a little bit long to explain (even if exciting) with details so I give you the conclusion : methodology to collect data in one side, and to standardize in the same language (english) on the other side, sucks like hell and the fact was, you can have a lot of dataā€¦which are so BAD that use it later for some stat analyses looks like pure waste of time. I did it for 3 rounds (6 years) and it was always the same methodological nightmare soā€¦Data are fā€¦ freaky and what people do with them as well (sometimes)

Iā€™ll be curious to read some other (good or freaky) stories with dataā€¦if someone want to share her/his journey withā€¦

1 Like

Yeah, something like this. Just something to play around with while learning Phoenix; nothing significant.

Phoenix is based on the language Elixir and I donā€™t think many people know about it here. It has some realtime features that I want to try out hence the reason I want to build a realtime based app.


Donā€™t know if this could help but it looks updated regularly

1 Like

@Groucha and @wanzulfikri I would surprise if you can find real-time data about covid. Reports depend on country/county base everywhere in the world and are usually provided once per day. The data is also no comparable in many aspects so it should be taken cautiously. Data collection differs by country or even per hospitals. The most you can do is collect the data as it is and select what to show.

Worst, it can differ from one day to the other without warning (eg. a country doing more tests, or a re-evaluation of what to include as cases).

In general every new report would replace the old one, and the old data will be deleted. It is considered for public info only, so many reporting institutions are not providing APIs or something like that. Past data is gone or hard to find.

The websites that are specialised in showing covid data are keeping a database of every daily report and part of the data. You have to find a daily feed or a daily updated dataset. The one shared by @Groucha is ok. I found one specifically for the Netherlands and used for a codepen (https://codepen.io/ecccs/pen/pojJyOd). I might work on it a bit more later.

I actually suggest you to try to look some of the current websites and see if you can improve their views - they are usually AWFUL.

There is one person who deared to make a cluster out of some international values, just for fun. In general, finding creamy info right now is not easy and even impossible. In some cases important data might not be collected. In other cases some collected data might be considered more confidential or socially and even politically sensitive, so is not shared. I wouldnt doubt parts of collected data (when collected) might be also monopolised: at the end, info has an intrinsic value.

@Groucha: We worked for the same sector!!! I was the project coordinator though (the person that revised that everything was ok and followed the fieldwork). Big projects too, with a lot of data analysis and data handling (very hand-on and technical). Same here: almost 24/7 job. A lot of money involved. But that was long time ago now.

1 Like

Itā€™s my biggest interest in this field haha ! Iā€™m really curious about how countries do their counts and how they communicate :slight_smile:
So much methodological issues + idem for tracking with ethical issues to add

How did you feel this job ? Could you give one or two examples of ā€œdifficultiesā€ you met in it ? I think I could die if I have to do, things are never ok xD