Refactor for changes in upstream 24 March 2020

datasets / covid-19

Novel Coronavirus 2019 time series data on cases

https://datahub.io/core/covid-19

1.16k stars 605 forks source link

Refactor for changes in upstream 24 March 2020 #36

Open rufuspollock opened 4 years ago

rufuspollock commented 4 years ago

There are changes upstream we need to handle (cf #33): https://github.com/CSSEGISandData/COVID-19/issues/1250

[x] US specific data with info by @anuveyatsu are we now pulling the US data into a separate file? If not, that would be useful i think e.g. us.csv
[ ] Also what other new data files or changes do we need to do?

rufuspollock commented 4 years ago

cc @westmich

westmich commented 4 years ago

@rufuspollock Thanks for tagging me on this. Any thoughts on the why? Are they having trouble keeping up with cases or is this data getting locked down?

anuveyatsu commented 4 years ago

One of the major changes is the removal of recovered cases (#39). On this, I'm planning to remove Recovered field from the data files (we may move it to archive directory if anybody needs it). WDYT?

berndblasius commented 4 years ago

Can you please leave the Recovered field in for the other countries which provide this info? I know there are some quality issues about recovered data, but it is crucial to calculate the number of active cases.

anuveyatsu commented 4 years ago

@berndblasius we depend on the upstream here so we can't keep it I'm afraid..

berndblasius commented 4 years ago

Hi,

thanks for the info. But I don't really understand the problem, for most countries confirmed cases are given by Johns Hopkins University in the daily reports. Or do I miss something?

Best Bernd

On Thu, Mar 26, 2020 at 2:18 PM Anuar Ustayev notifications@github.com wrote:

@berndblasius we depend on the upstream here so we can't keep it I'm afraid..

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

anuveyatsu commented 4 years ago

@berndblasius they have announced that they will stop providing the recovered data sometimes soon

rufuspollock commented 4 years ago

@anuveyatsu one option is for us to write out the current recovered data to a separate file here and then integrate when we do the processing so it stays around and we don't have a breaking change?