Following @HenkPoley's comment I've fixed the logical mistake with the ICU calculation (fractions are hard), updated the constant used, and also updated the method of calculation (to be a ratio of the reported active cases and not a ratio of the estimated cases). The table looks considerably redder and worrisome now :(

In this notebook I compare the number and change of confirmed cases per day with the number and change of deaths per country.

Output: https://caglorithm.github.io/covid19-analysis/covid-cases-to-deaths/

• The notebook follows ideas from Tomas Pueyo's article "Coronavirus: Why You Must Act Now"
• Estimated time of infection is 20 days before death
• Fatality rate of each country is calculated for estimated infections
• An overview of the number of confirmed cases vs. the number of deaths per country is given
• Distance of the maximum of cumulative deaths and new deaths to the number of infections is used to estimate the progression of the infection across countries.

This adds a per country overview and projection of immunisation progress, ICU need and other data.

Details:

• Data: progress projections is taking the overview table and adds:
  • Quantities per 100k population (by adding populations data),
  • Estimation of total cases from deaths,
  • Current immunisation percentage,
  • Current need for ICU beds per 100k population, and
  • Projection into future of immunisation percentages and ICU needs.
• Presentation:
  • Color-bars-table for current and future immunisation percentage by country - how far do we still have to go until this is over.
  • Color-bars-table current and future need for ICU beds per 100k - how bad is the medical crisis currently and in the future (unless growth rate changes) in each country.
  • Full overview table with estimation and projections for all countries and the base quantities (cases, deaths, etc).

Changes:

• Refactor overview helper code. Only create helper module (without editing the original overview notebook).
• Add overview template and country mappings into _data folder in the repo (to reduce external dependencies).
• Add world population data file.
• Add thumbnail for progress projections dashboard.

I was experimenting with using jupyter notebook, pandas (data transformations), jinja2 (html, visual) to create dashboards with fastpages. Bulk of visual heavy-lifting is in jinja2 and data with pandas.

Is the something you'd want to add?

Trying to reopen https://github.com/github/covid19-dashboard/pull/232. For some reason this PR is only for one recent commit, I couldn't open a PR with the reverted commits (revered https://github.com/github/covid19-dashboard/pull/245). Not sure how to reopen with the reverted commits.

Currently all notebooks and all images are committed every deployment (every 6 hours). This has some big advantages:

• Reproducibility.
• Simplicity.
• NB viewer for PRs.

This his also many disadvantages:

• .git folder data is getting very large (900GB).
• Sync times are getting slow.
• Notebook files changes are hard to review (not diff-friendly).
• NB-viewer is not really a reliable way to render the notebooks (with all the various JS and HTML used in them),
• Bot commits are adding lots of noise to the commit history.
• Changes need to always be rebased on top of bot commits even if no actual changes besides deployment happened.

Fortunately, there is a tool that makes life with Jupyter notebooks and git much easier (at least for me): https://github.com/mwouts/jupytext. It is an incredibly comfortable way of working in both an IDE and a notebook simultaneously, but most importantly it makes git life much easier:

• Notebook files are "paired" to .py files (with some comments to store structure and markdown) and are synced together locally.
• .py are committed to git, whereas .ipynb files are not (e.g. ignored in .gitignore).
• During deployments a command to generate .ipynb is used (they will contain only inputs, and no outputs), and than papermill can run them to fill them with up to date outputs. Nothing needs to be committed, because .py files haven't changed.
• For local development Jupyter will automatically open the .py files with the metadata as notebooks when jupytext is installed. Otherwise a short Makefile target can create / update all notebooks using the same script that's used during deployment.

In my own repo (where I work on the various visualisations / calculations) I've switched to this system and am very happy about it:

• Only meaningful commits are in the history and there is no tradeoff between deployment schedule and git noise (although I keep it to once a day, since there is no hurry).
• I can clearly see changes I'm committing using the diff on the .py files.
• The changes are all very light weight since they don't contain data and random JSON noise.

With COVID-19 not likely to be over quickly any time soon it might be a good idea to give this a chance. It's pretty simple, and can be gradual (notebooks can be moved to this system and git-removed gradually).

Hi, I added an overview notebook for Australia showing state-wise data of COVID-19. Also:

• I added recovered-cases data in the overview page.
• I tried to make the design better -- clean and modern. I did it for Australia, and the plan is to update the design for rest of the overview pages after getting your feedback.

You can check the original repo and the new design in the screenshot below.

Let me know if you find the changes useful. I'm happy to update per your feedback.

Thank you, Ash

Looks like some of the changes could be breaking changes

See: https://github.com/CSSEGISandData/COVID-19/issues/1250#issue-585788363

The procedure for handling this is any failing dashboards will be hidden from view from the home page if notebooks happen to fail. Just want to give everyone a heads up

@twiecki @jwrichar @jbduarte @pratapvardhan @aimran-adroll @avyfain

Some time-series charts plotting estimated daily values for effective reproduction number (covid-19) for Canadian provinces. I welcome questions and comments on it.

This PR will create top links between pages

CHANGES

• covid19-overview.ipynb will remain like before
• covid19-overview-us.ipynb and covid19-overview-europe.ipynb will use covid_overview.py
• covid19-overview.ipynb leaving it as is, for giving notebook experience
• Only covid19-overview.ipynb will be visible on homepage.

COVID-19 Growth Rate Prediction | COVID-19 Dashboards

Predictions of COVID-19 Growth Rates Using Bayesian Modeling

https://covid19dashboards.com/growth-bayes/