search

CSSEGISandData / COVID-19

Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE
https://systems.jhu.edu/research/public-health/ncov/
29.1k stars 18.39k forks source link

Reconciliation summary report useful? #1959

Open judepayne opened 4 years ago

judepayne commented 4 years ago

Hi, is it useful anyone to post a daily reconciliation report (country level only) like the below here on a daily basis just after JHU have updated their data? I use it as a quick way to see that the data looks good for today or if there is a country (or two) which appears to be out and I need to put in a patched value (or two) for the graphs I am generating on top. report . . .

Reconciliation summary report

04-Apr-20_01:20 UTC +1

Filter params:

cases asbolute threshold:  20
deaths absolute threshold: 10
cases percent threshold:   2
deaths percent threshold:  2

(Any differences below these thresholds have been filtered out.)

In the tables below the (absolute) dif and percentage difference are jhu vs. the highest of the other sources. Negative diff/ % diff means jhu is behind.

country jhu wikipedia worldometer diff % diff
Panama 1,475 1,475 1,673 -198 -13.42
Argentina 1,265 1,265 1,353 -88 -6.96
Czech Republic 4,091 4,091 4,190 -99 -2.42
Australia 5,330 5,358 5,454 -124 -2.33
Denmark 3,946 3,757 3,757 189 4.79

Deaths

country jhu wikipedia worldometer diff % diff
Canada 179 178 208 -29 -16.2
United States 7,087 7,381 7,391 -304 -4.29
Switzerland 591 604 591 -13 -2.2

Countries with no data in the jhu dataset:

Saint Vincent and the Grenadines Burma Holy See West Bank and Gaza Sint Eustatius and Saba"

Full Report

cipriancraciun commented 4 years ago

@judepayne Nice reports!

One question though, how do you collect the data for WikiPedia and WorldOfMeters? Are you entering them by hand? (In case of WikiPedia, you might screen-scrape-it, but for the other one?)

(I'm asking because I've built a derived dataset based on JHU, NY Times and ECDC and I always see these differences. I've described the derived dataset in #1281 and it is available at https://github.com/cipriancraciun/covid19-datasets . Moreover if scraping WikiPedia is easy, I might also add that to my dataset.)