aatishb / covidtrends Goto Github PK

Similar to how when the USA states appear as the selection choices when the "US" region is chosen, would it be difficult to add another level of selection for each state and have the counties appear as choices for that state? The raw data for that appears to be in the NYTimes data set.

Thanks for the excellent work on this incredibly helpful and open site.

Displaying summary on the top or bottom of the home page.

It would be nice to see the following numbers:

• Total Confirmed cases (% increase/decrease)
• Total Active cases (%)
• Total Recovered (with %)
• Total Deceased (with %)

Fantastic project, super useful. Thanks @aatishb 👌🏻

While we can view the data of all countries superimposed, it would be valuable to have a curve representing worldwide totals.

This would only involve adding a "World" entry to the countries list whose data is the sum of all other countries. This could be done client-side.

Hi,
Great tool. Is it possible to add The Gambia in the list of the countries?

Thanks in advance.

I know that the data resolution is 1 day, but the animation would look much better if a higher framerate could be achieved by interpolating.

Hi there,

could you show how old the dataset is and the estimated time the new dataset will arrive?

That would be nice indeed.

Am considering revisiting a few decisions, triggered by reading #47, and a few emails from folks wondering why their countries aren't included in the deaths graph. At the moment this is mostly a note to self, however let me know if anyone has feedback on this or wants to take a stab at implementing this.

Initially I'd set minCasesInCountry to 50 as the global confirmed cases data was noisier below this value (the variable refers to the threshold for including a country on the graph). However I think we could have different thresholds for minCasesInCountry and minDeathsInCountry. Just looking at it by eye, it seems like setting minDeathsInCountry to 10 creates a clean looking graph.

Also when looking at US states, it seems like both can be set to 10 just fine as the case data has less variability for low values (< 50).

Finally whenever minCasesInCountry is 50, the minimum on the visual settings should probably be set accordingly, to emphasize this threshold.

Some countries have more reliable case number reporting and after the total number of cases gets big enough the rate stabilises itself and a week might be too pessimist for such countries (and might be hiding the effects of the measured taken).

A parameter to change the size of the rolling horizon would be interesting.

Thanks slot for this tool! It's ingenious.

Can you please add Europe selection tab, for quick select Europe countries? :)

While the log vs log plot is great for comparing the shape of the trajectory of individual countries, wouldn't it make sense to be able to normalize by the total population, so it becomes possible to compare between countries more accurately?

As the pandemic goes on, the earliest numbers of cases have less and less relevance to the current pool of infected and the spread of the disease. I think it makes sense to replace the current lifetime total of the X-axis with a moving window of 8 weeks. Perhaps 12 weeks to be even more conservative. Thanks for the great visualization and code!

Hi.

Why not using number of cases per capita ?

Great tool! Thanks.

It would be great to have additional aggregates of countries in their respective economic/political blocks & institutions, e.g. EU, Schengen Area, G7, G8, G20, Emerging Markets, so you could (for example) compare a single EU country to the EU total… and you'd have a possible indicator for economic recovery, when e.g. both the G7 & the emerging markets display a downward trend.

PS: and a world total would also be a good thing to have.

Would it be possible to choose to view the derivative of currently drawn curve instead? It could be better visualization of the divergence from the exponential path...

A small nit #/#/## is ambiguous as to which is month and which is day.
Less ambiguous e.g. "4 Apr 2020?
Or a localized format (not sure if this is possible)?

As the slider moves, the day number becomes clear, but still requires a mental switch for non USA people.

Hi all, just wanted to share an update.

@mustpax and I are working on refactoring the graph component. Currently the graph component uses many data watchers to watch for changes in the dropdown menus on the page and react accordingly.

This design has led to some inconsistent behavior (such as race conditions with multiple watchers firing) that came up with a few PRs (e.g. #76 most recently). Rather than manually come up with one-off fixes for these, we're hoping to simplify the logic of the graph component to avoid this problem altogether. This should also make it a bit easier to fork this and add data sources / update the UI, etc.

I think this shouldn't clash with most existing PRs, as most of your code changes have been in the root vue layer rather than in the graph component.

When I select "Reported deaths" there's much less countries than for the "Confirmed cases" dataset, even countries with more than 5 deaths now.

Would you please update countries the list? Thanks.

This is different the the existing active cases suggestion I saw. This would be another member in customise - "confirmed cases", "deaths", "active cases".

The goal would be to gain a better visual understanding of the trends which show when hospitals are unable to handle the number of active cases.

For some analysis, it will be useful if evolution is shown from DAY 0, day of first case.
Great job.

Hello,
thank you for your your work. I'm very happy to find clear visualization that helps to understand what is going on in all this flood of information.
Under what license you publish your work on https://aatishb.com/covidtrends/ ? Do you agree to publish snapshots of your visualization on wikipedia / wikimedia and related websites with full credits to you and other contributors?

Can we mark the date of country-wide lockdowns on each line to help visualize how long after lockdown the slope starts to change. Anecdotally looks like some 3 weeks after lockdown for Italy/Spain. The lag presumably depends on the level of lockdown compliance.

Hi, Would really be useful to be able to select the states and provinces of US & Canada so that they could be compared against each other!

Problem: An effective way to reduce impact of this virus is to communicate best practices faster than the virus infects. Therefore working short links are critical.

When clicking on "Create shareable link" I'd copy (for example) https://aatishb.com/covidtrends/?location=Hong+Kong&location=Italy&location=Singapore&location=Spain&location=Taiwan
Now when pasting on certain websites, specifically facebook, once posted the clickable link results in : https://aatishb.com/covidtrends/?location=Hong+Kong with the rest truncated.

This issue highlights two problems. lengthy links and two the links truncate

This makes it more difficult to communicate quickly when discussing different strategies and the efficacy thereof.

Potential Solution: ideally the link is short thus doesn't demand a lot of space in tweets and cluttering communications. I know you can use a url shortener, but I personally hate those and will very rarely click on them.

Hi all,
Allow me first of all to thank you warmly for this remarkable work.

I got driven here from the video : How To Tell If We're Beating COVID-19 you did with minutephysics.

By watching related video on the subject, I realised a important in information values seems to be hidden in the background (the Growth factor behaviour). Thus I would love to be able to see how the Growth factor (and derivative) is behaving in time. I think it could be super useful to guess when states arereaching their inflection point.

3Blue1Brown :Exponential growth and epidemics

[EDIT:] My bad, I did not understood that "New confirmed case" axis is actually the growth rate. Maybe a small addition in the label might ease the understanding. Well if derivative could be plotted I'll be delighted.

All the best

Sincerely

In this graph (https://aatishb.com/covidtrends/), can you add lines which indicate how many days it takes to double the cases? For eg, the line US is following indicates doubling every 5-6 days. A parallel line below it would indicate, say, 10 day doubling.
Regards,
Mandar

Since the US is such a big country, is there a way to make the graph show the stats of individual states (ex: Arizona, Texas, California, etc.)?

I believe, ploting active cases instead of total confirmed cases would allow better comparison of effective measures against spread of covid-19 in different countries.

Because recovered people are not infectious, the current graph is biased towards countries with a greater fraction of healed (and died) people among total cases, like China and Korea. Bigger amount of closed cases actually makes the denominator bigger, but it is not government measure for containment, making countries that first got it look better than they actually are.

Working with active cases would also help with the second waves, otherwise invisible on the current chart.

Given the axes, logarithmic growth is characterised by a straight line with gradient 1.

Whilst you can almost see this from the chart, it would be helpful to have this y=x line as a reference - so you can see how far a country has deviated below logarithmic growth.

I should say, I think this is an absolutely excellent viz. The best I have seen out there for measuring progress

It would be great to be able to compare Hubei, New York and Itally. This would combine different sources of data but give a better comparison than just all of China compared to all of the US to just Italy.

each states should have different curves

Hi I've taken a brief first look at the markup and noticed a few issues and low hanging fruits in regards of html5 validity. I could create a quick pull request if you want to. The minor details I've noticed:

• remove the two meta content lines

  <meta content="text/html;charset=utf-8" http-equiv="Content-Type">
  <meta content="utf-8" http-equiv="encoding">

and replace that with the best practice of

<meta charset="utf-8">

and position it right before the title element (see https://htmlhead.dev)

• move the twitter card as well as the open graph meta data to the head element
• remove the two closing tags of img elements for the two logos

best regards ralf

If the data were available, could this same method be applied to other similar diseases? What information could we derive from the comparisons of historical diseases to Covid-19? What would be the implications of similar graphs and likewise, what would be the implications of wildly different graphs?

Are you aware that Vue.js is running in development mode on https://aatishb.com/covidtrends/ ?

There's lots more info about this here: https://vuejs.org/v2/guide/deployment.html

Your graphing tool is a really great resource, thank you!

I would love to have an option to see the data scaled per-capita instead of absolute numbers. This shouldn't change the trends (right?), but it would give a better sense of relative scale between countries. For example, the US is currently beating Italy (as of 3/27/2020) on an absolute scale -- with the US at ~70k weekly and ~83k total and Italy at ~40k weekly and ~80k total). However if you adjust for the populations of the countries, you see these figures (trivial math -- try it yourself):

US total: 25.6 cases per 100,000 people
US weekly: 21.4 cases per 100,000 people
Italy total: 133.2 cases per 100,000 people
Italy weekly: 65.4 cases per 100,000 people

This paints an interesting picture. I suppose it shows that right now, even though the US is starting to beat Italy in terms of absolute quantities, Italy is still, for the moment, being hit proportionally harder. I wonder if/where the point could be that the US passes Italy up in that metric.

As both Hong Kong and Macau are separate legal jurisdictions and travel between them and mainland China require passports. Covid19 cases should not be consolidated as one for the three territories. China should be mainland-only (exclude HK & MO) with Hong Kong and Macau being their own clickable entities.

Every day Ontario Gov't publishes at 10.30 am their stats in detail. As of yesterday 9/4 there were 6237 cases open. Your chart shows 5276 but I also note the chart title says 08/4. Then again I look at Quebec data same chart and we we see 10031. According to the Quebec Govt web site as of 6pm April 8th they had 10,912 cases. Since these are publically available data sets - I am wondering whether your data is behind for ontario and then even so whether its accurate for Quebec?

Could you please add a chart for comparison with Influenza please to reduce panic among people that this is a killer virus (capable of wiping out humanity from the face of the earth if left unchecked)?

https://www.hopkinsmedicine.org/health/conditions-and-diseases/coronavirus/coronavirus-disease-2019-vs-the-flu

Some countries get new outbreaks after recovery. Are there already ideas how to visualise that? Or would it be good to discuss that here?

I really like this representation, as it's one of the few really clear ones. Thanks for making it!

Hello! Thank you for this wonderful visualization. I was wondering if it would be possible to combine your interpretation of data along with a "Community Mobility Report" (such as this one from Google https://www.google.com/covid19/mobility/) to see if there are any correlations between the decrease in mobility and the spread of the virus. If this could be broken down by country/state/region that would be even better!

Been getting some emails and messages that when we push an update, it breaks the site on Microsoft Edge. The issue seems to be that Edge is holding on to the old cache. Is there any way we can force the browser to reset cache after pushing a change?

Here's what Edge users see when this happens, so this is not a graceful break.

According to JHU data and WHO data, Hongkong should be list as a state-level in China.

At the moment the two css files as well as the definitions file for vue aren't minified.

Hi,
It would be great to have state-wise data for India similar to US, China, Australia etc. Data from covid19india team's API might help.

API: https://github.com/covid19india/api
Dashboard: https://www.covid19india.org/ (Github: https://github.com/covid19india/covid19india-react)

Thanks

Somehow, the current Data Point for France has less total cases than before. This is visible as a backward bend in the graph connected over time.

Can we have each country with different color while plotting and making progression in graph? It will be really useful to differentiate in animation.

Could you include a Cases/Death per 100,000 People statistics of the Countries. This would introduce a scale not dependent on the population of a country.

Brazil is a big country with many different hotspots. Including a breakdown per state (as per US, Canada, Australia) would help visualize what's happening. This should be fairly easy from the Health Ministry data (https://covid.saude.gov.br/)

Is it possible to get the page translated? It's a very useful tool to explain this event in an easy to understand to people that doesn't have access to English.
I could help on the translations (at least to Spanish).

Good job. Thanks for sharing.