Skip to content

Latest commit

 

History

History
232 lines (146 loc) · 14.6 KB

README.md

File metadata and controls

232 lines (146 loc) · 14.6 KB
Raw

Ultimate post on Coronavirus for @opendatascience telegram channel

Contents:

  1. Learning and enlightning others
  2. WHO's advice for the public
  3. What you can do about it
  4. Remote work

Dear readers and subscribers, we would have ignored coronavirus (COVID-2019) theme completely, wasn't it for 64% of you who asked for at least some information to be posted.

So we came up with rather big and thorough post, available as a GitHub page, so you can submit your pull requests with updates / additions to the information provided.

We followed some rules, which we find adequate and which are based on the common sense and hope that all the contributors will comply with them:

  1. Provided information should be correct, better if it is verifiable.
  2. Source should be provided not to spread fake news data.
  3. Biases and distributions should be taken into account: raw information is not that representative and can misguide opinions.
  4. If applicable, information should be actionable — readers should get a clear picture of what they can do after reading it, not just get upset or worried.

Learning and enlightening others

Briefing:

  • Coronavirus is somehow similar to common cold: spreads throw air or surfaces, causes dry cough and pneumonia on late stages.
  • Despite being similar, it is much more contagious, because people are not even partially immune (yet).
  • Danger comes from late stages and overloaded healthcare system.
  • Diagnostics is hard, requires special tests.
  • Most threatening is exponential (or viral) growth. Each carrier is able to infect ~2.5 (source) people.
  • More dangerous to elderly people.
  • There are 2 strategies exploited by governments: letting build 'herd immunity' by quarantining older people and restraining all population, to flatten the curve to fight healthcare system overload. The concept of curve flattening is criticized due to possible overestimating the capacity of the healthcare system.
  • Other danger comes from panic and spread of misinformation (or censorship of that information by government): check for the fake news and trustworthiness of the source.
  • There is probability that the humanity will be infected and build immunity.
  • Maps and sources of spread often do not take into account availability of tests (why it matters) thus richer countries tend to have more diagnosed cases, thus, GDP correlation to cases per capita matters with 0.6 for Kendall and 0.8 for Spearman (Post in russian, data sources).

WHO's advice for the public

WHO stands for World Health Organization and that's one organization you can trust in terms of health recommendations (it's like Open Data Science, but about Health 😉).

  • Wash your hands better: use alcohol-based liquid or soap, wash your hands deliberately, like surgeons in the movies do.
  • Wash your hands more frequently: adopt a habit of washing hands before having a meal and after coming home / to work.
  • Maintain social distancing: stay away from people sneezing or coughing.
  • Avoid touching your face, nose, mouth and eyes.
  • Practise respiratory hygiene: cover your mouth and nose with your bent elbow or tissue when you cough or sneeze. WHO Source
  • Common thin face masks are not effective to protect from getting infected (WHO).

Articles and sites covering basics on coronavirus

What happens next

Interactive showcases: will there be the second wave of infection and what happens in the nearest future.

Link: https://ncase.me/covid-19/

Why exponential growth is so dangerous

Interactive article on how virus spreads

Kevin Simler's outbreak interactive demo shows and teaches from the case of single person how virus spreads.

Project website Link: Github, Node JS

Alternative look on fatality and mortality rate

English translation of article acquiring 1 mln views in russian.

Link: Medium

Washington Post's article with demo

Link: Washington Post

Visualization 'How the Virus Got Out' by The New York Times

Link: NYTimes

Korean cluster's Reuter's infographics

Link: Reuter's site

DmitrySerg's spread simulation project

Link: Project Github

📹 Videos

Pornhub's insights on traffic

Link: Page

What you can do about it

  1. Follow WHO's advice to lower your chances of getting infecting.
  2. Stay inside, switch to remote work if possible.
  3. Spread the word about the pandemia, share trustworthy information.

What you can do (as data scientists and active community members):

Most of the information about COVID-2019 causes panic and just leaves you with anxiety. But there are some projects, in which any dear data scientist reader can take part and actually do something about the virus (and that's worth repeating), apart from following advices to lower spread of virus and informing others with reliable sources.

When Will COVID-19 End (Data-Driven Estimation of End Dates)

Independent research focused on predicting end of the pandemic.

Link: https://ddi.sutd.edu.sg/

Welltory's Wearables Open Data Research

Popular description: ProductHunt Github: repo

Github's blog post on open collaboration projects

Lot's of initiatives including those not covered by current post.

Link: Github's Blog

Flattenthecurve

Ultimate website is also community-driven, you can help in any way: verifiyng information or translating the content. Despite being named after controversial concept, it is very good entry point to get information about COVID-19 (US specific information mostly but anyway!)

Website: Link. Contribution guidelines: github.

folding@home coronavirus-specific projects

What: Donate your GPU power to help researchers develop drug against COVID-2019

Link: Discussion on Reddit

or you can do it with

Browse Scientific Articles about Covid-19 & SARS-CoV-2 with SciBERT-NLI

Covid-19 Browser is an interactive experimental tool leveraging a SOTA language model to search relevant content inside the COVID-19 Open Research Dataset (CORD-19) recently published by the White House and its research partners.

Link: Github

boinc project on rosseta@home

Link: Post on Ipd and discussion thread

CLI-tool for COVID related data retrieving and datasets

Unified transcoding tool and cli for covid-19 related national and international datasets. You can help with the development or use it for you research.

Link: Github

Building CT scan pneumonia detector

What: CT scan / Xray pneumonia detector.

Note: CT scan detection might be ineffective for COVID detection, because pneumonia is rather late stage on which intensive care is already required.

Project 1: Github Project 2: Github

There is some datasets and competitions about COVID-19 on Kaggle

You can do some competitions on kaggle

All related itmes on kaggle by covid19 tag

If you did some work related on COVID-19, don't be shy to use opportunity: get free tools

COVID-19 Simulations

These simulations/modeling/scenario-based-forecasting projects include various methods and techniques; some are more accurate in terms of the implemented epidemiological model in use, whilst others offer faster modeling at the expense of standardization and in return interpreted result soundness and integrity.

Remote work

Some of us will spend more time working remotely. We all have different backgrounds and trivial information might be valuable for people just starting to work remotely. Remote work due to the quarantine or self-isolation may seem challenging at first, but there is nothing to be afraid of. We collected some articles and FAQs which will help to transit to remote mode.

Links:

Dashboards, Maps, Aggregations, Noteworthy articles

More links