search

CSSEGISandData / COVID-19

Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE
https://systems.jhu.edu/research/public-health/ncov/
29.13k stars 18.43k forks source link

There is No Reason to continue developing with this dataset #1485

Open sibblegp opened 4 years ago

sibblegp commented 4 years ago

The maintainers have shown a clear disregard for the thousand of developers, sites, and visitors to those sites. Broken or missing data and breaking changes have ruined the ability to reliably depend on these numbers.

Abandon ship people. Where can we find/make something better? I'm happy to build and host an API as long as we can find an accurate data set that doesn't keep breaking.

joaopedrosantos commented 4 years ago

You can start your own

rben01 commented 4 years ago

I can't vouch for the accuracy, timeliness, availability, or future consistency of this dataset, but at least it is in a useful format: https://coronadatascraper.com/#timeseries.csv

regattaguru commented 4 years ago

Management of this data set has left it with zero credibility.

ulklc commented 4 years ago

This repo have an valuable dataset. You can check alternatively: https://github.com/ulklc/covid19-timeseries

Hikmer commented 4 years ago

Its absurd this fell apart in the moment of crisis....I mean aren't we all at home??? Possibly the biggest fail of all time in terms of data practices.

rben01 commented 4 years ago

This repo have an valuable dataset. You can check alternatively: https://github.com/ulklc/covid19-timeseries

Unfortunately that doesn't seem to have data for US states

jasonbrown1965 commented 4 years ago

This repo have an valuable dataset. You can check alternatively: https://github.com/ulklc/covid19-timeseries

Unfortunately that doesn't seem to have data for US states

John Hopkins is on the list?

dereklarson commented 4 years ago

This repo seems to do a decent job cleaning up the data from the John's Hopkins data available here, though it's just for the countries (e.g. not states):

https://github.com/datasets/covid-19

hippodribble commented 4 years ago

I put some Python code at #1458 to transform new files to the older format. Maybe it will help.

amnonbc commented 4 years ago

I came here after the WHO stopped daily publishing of their data. I have not found any better data source.

ghost commented 4 years ago

@rben01 posted https://coronadatascraper.com/#timeseries.csv earlier and they're starting to cross-check with NYT. I learned about it on one of the covid-atlas.slack.com channels. Not to discount the efforts of this project in any way, but coronadatascraper.com seems more disciplined.