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

CSV INPUT PROBLEM? Confirmed Cases For Every Country Set to Zero again.. #747

Open DavidODonovan opened 4 years ago

DavidODonovan commented 4 years ago

For the second day in a row all confirmed cases for all countries has been set back to ZERO. I just glanced at the raw .csv files and there is a column date value but on every row there is no entry in that last column. Seems like you have a problem with the input mechanism that is building the .csv file. If that is an online dashboard of some sort then there is a problem with your code.

Waldo79 commented 4 years ago

The past week or so I wait for at least 4 hours after the first post to download anything (don't waste as much time downloading corrections.) The last couple of days with the US so screwed up to the point it's useless for a time history, I quit downloading at all.

Chris-JAC commented 4 years ago

First - really thankful for what these guys are doing. I think it is only 3 people handling this so mammoth job. So really thankful and trying to just help here.

I note 2 problems with the .CSV files.

  1. Empty last column. Same as yesterday.
  2. The line-end is not consistent. Yesterday was CR-LF, today is back to LF. I'd prefer CR-LF as that means it imports nicely into Excel, but if LF then I can rewrite my macros again.

No need to reply of course. You guys must be getting thousands of comments like this to deal with (or not). Thanks again. keep it up.

Bost commented 4 years ago

I think it is only 3 people handling this so mammoth job.

:man_facepalming: So if these guys get sick... this repo has 8.6K stars and 3.2K forks and there are important decisions being made around the world based on the data from this repo. Literally life vs. death decisions... 3 guys and all of them sitting in one office I guess.

This is reckless and irresponsible.

peterdrier commented 4 years ago

check the daily files folks, not the pivot. The data is better.

JiPiBi commented 4 years ago

@peterdrier There are also some issues in the daily values : lines in double , date formats for the first files

IMO the only difference I see is that they broadcast the new daily file with only prepared and not filled pivot tables , and you got the updated tables some hours after , dont know when because i'm in bed ...

Now in my jupyter notebook , I test if the sum for the new day column is equal or not to zero before doing the rest of the calculations , if not I stop