boogheta / coronavirus-countries Goto Github PK

suggestion by @MaximeSCPO

• build minvalues for perday only
• draw zero line on series
• disable logarithmic for perday (really close #25) => merge menu lin/log with cumul/daily
• handle negative stacked

Ideally a CSV file for a country with each row having at least a date, a region, and a list of cases numbers.

Possible sources:

• https://coronavirustechhandbook.com/data

• https://github.com/covid19-eu-zh/covid19-eu-data

• https://covid19-eu-data-cache.now.sh

• South Korea ?

• Iran ?

• Netherlands ?

• Sweden ?

cf https://twitter.com/Acolandre/status/1251975072665001988

For more readability

Also for logarithmic multiples view?

(as suggested by @jacomyma)

Some numbers appear multiple times - it rounds to larger numbers than the step.

Example:
https://boogheta.github.io/coronavirus-countries/#deceased&multiples&countries=Italy,Iran,Spain,France,South%20Korea,Germany

Question: what shall we do when multiples cases are selected?...

simplebar: like in https://ourworldindata.org/coronavirus (suggestion from @MaximeSCPO)
d3brushing: https://www.d3-graph-gallery.com/graph/stackedarea_template.html

• add slider
• display limits
• zoom in url
• fix overlap of handles value in d3-simple-slider (johnwalley/d3-simple-slider#119)

This tools is one of the best visualizations on the official case data out there. Congratulations 🎉

It would be really cool, if instead of the single values, one could also look at 7-day averages.
Main reason for this is, that relevant data (such as daily confirmed cases) are very noisy (no lab work on weekends, etc.).
This is why many analysts and even law-makers now consider the 7-day rolling average as the more relevant number for comparisons and to detect trends.

To avoid having the legend fill the screen with too many countries

• series/stacked legend
• small multiples legend

cf https://twitter.com/rebeccapbp/status/1241060148195528711

Hello,
Your dataset was added to CoronaWhy (https://www.coronawhy.org/) Data Lake on Dataverse as a piece of common COVID-19 data frame https://datasets.coronawhy.org/dataset.xhtml?persistentId=doi:10.5072/FK2/FW5FB6
Would you be willing to help with the maintenance of your dataset in Dataverse, e.g. adding the relevant metadata and keeping the dataset up-to-date? That will help to make the dataset findable and accessible for the medical science community.

I'm not completely convinced it is meaningless for new cases, but definitely for stacked

• add interface checkbox in cases title when in multiples
• handle proper fields:

all tested = {
   tested negative
 + confirmed = {
      recovered
    + sick = {
         confined
       + hospitalized
       + intensive_care
    }
    + deceased
  }
}

adjust cases legend?

• use different level strokes to display contours

TODO:

• Review menu titles to free some space for more switch buttons
• Add switch menu to display new cases instead of curves
• Plot it as histograms
• handle negative new cases on "currently sick"
• handle hover histograms (+ fix clearTooltip being called too much)
• Limit number of parallel countries possible in histogram series ? or force small multiples in new cases mode ?

cf https://twitter.com/cq94/status/1241684696955666433

TODO:

• Gather data for each country in a CSV with one line by country and one column per type of events, such as "closing schools & universities", "closing borders", "home quarantines"...
  -> https://www.acaps.org/covid19-government-measures-dataset
  or https://figshare.com/articles/CORD-19-policy_dataset/12055860/2
  https://covidamp.org/data?placeType=affected&type=policy
  https://www.who.int/emergencies/diseases/novel-coronavirus-2019/phsm
  https://github.com/OxCGRT/covid-policy-tracker
  option : reuse 100 based index from OWID cf low https://ourworldindata.org/coronavirus/country/france

• display bubbles on plots

cf https://twitter.com/MeringuemanKe/status/1244982782876622850
https://twitter.com/coulmont/status/1244654636662996999
https://twitter.com/aatishb/status/1241855558690516993

https://aatishb.com/covidtrends/

warning: log scale would be useful too, but maybe lin against log are not most useful and it would be acceptable to toggle only one log switch for both axis at once?

cf https://twitter.com/Dajjal2237/status/1262712154555383808

Wouldn't it be useful to be able to export a given view in SVG/PNG (with a white/transparent background ?) for further work with other tools ?

The Coronavirus Country Comparator is my goto solution to really understand what's happening in the world. Longterm trends are best visible in this app.

Therefore it would be great to also be able to look at "vaccination" graph (if such data can be obtained easily).

Ideally, this could be combined with an additional "immunity" graph, which takes both vaccination and former infections into account.

In any way: Thanks for this tool!

@boogheta Thanks for this excellent site. I keep tracking a few countries, including France. A recent adjustment of their numbers gives a significantly negative confirmed case for May 20, 2021. Would it be possible to always set the vertical axis lower bound to zero, to hide such anomalies?

As suggested by @bjperson and implemented by @njoyard here https://corona.njoyard.fr

• add buttons to interface
• handle legend too big (depends #30)
• handle url too long : option all with list of unselected ? unnecessary
• handle small multiples : limit number of columns to a minwidth and start scrolling the div if more? unnecessary

Source data possibly needs to be updated - I noticed that UK deaths numbers are identical for the last two days.

Readme.md at https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_time_series states:

---DEPRICATED WARNING---
The files below will no longer be updated. With the release of the new data structure, we are updating our time series tables to reflect these changes. Please reference time_series_covid19_confirmed_global.csv and time_series_covid19_deaths_global.csv for the latest time series data.

time_series_19-covid-Confirmed.csv
time_series_19-covid-Deaths.csv
time_series_19-covid-Recovered.csv

Figures in time_series_covid19_deaths_global.csv are correct (as per readme warning).

Suggestions from @paulgirard:

• display small dots on curves by day (and tooltips with value on them?)
• hover curves should be a line on dots rather than a band
• in series mode, use radio buttons for cases instead of switches ?

when you switch from 'series' to 'stack' the order of colors doesn't match the countries.

Add in menu : since 10th case for deceased and since 100th case for confirmed

Currently it displays the total world value, which makes sense when comparing multiple countries, but would make more sense when there's only one

Attach the color at startup with ordered countries by confirmer or deceased

Known data issues:

https://github.com/CSSEGISandData/COVID-19/issues

• US recovered: CSSEGISandData/COVID-19#1113
• France recovered: CSSEGISandData/COVID-19#1169
• plateaux France, Spain, CSSEGISandData/COVID-19#1187

fix https://github.com/CSSEGISandData/COVID-19/pull/1237/files

Alternatives:

• https://coronadatascraper.com/ ?

Functionality available here ("normalize to population") : https://covid-stats.theo.io/dashboard_draggable.php

PS: Thank you for your work ! ;-)

Running the consolidate_uk.py file is returning the following error:

$ ./bin/consolidate_uk.py > data/uk.csv
Traceback (most recent call last):
  File "./bin/consolidate_uk.py", line 55, in <module>
    print(",".join([d, b, str(countries[b]["confirmed"][i]), str(countries[b]["deceased"][i])]))
IndexError: list index out of range

at least for World

JHU's data is adding new provinces and countries when new ones are reported, so the list will need to be completed from time to time @robindemourat
Here are the few missing as of tonight:

WARNING: missing population for Canada / Northwest Territories
WARNING: missing population for Canada / Recovered
WARNING: missing population for Canada / Yukon
WARNING: missing population for World / Saint Kitts and Nevis
WARNING: missing population for World / Mali
WARNING: missing population for World / Guinea-Bissau
WARNING: missing population for World / Kosovo
WARNING: missing population for World / West Bank and Gaza
WARNING: missing population for World / MS Zaandam
WARNING: missing population for World / Burma

https://docs.google.com/spreadsheets/d/1e703pe3GmBQt0i2yAOS0F6Bhxy91U1-NTB6JMRSTzc0/edit?usp=sharing

Originally posted by @boogheta in #1 (comment)

whole scope gets at the bottom of the list weirdly

Sources : https://coronavirustechhandbook.com/data
https://github.com/covid19-eu-zh/covid19-eu-data
https://covid19-eu-data-cache.now.sh

• update help text

• France
  https://github.com/cedricguadalupe/FRANCE-COVID-19
  https://www.data.gouv.fr/fr/organizations/sante-publique-france/
  https://www.data.gouv.fr/fr/datasets/5e74ecf52eb7514f2d3b8845/
  https://www.data.gouv.fr/fr/datasets/5e7e104ace2080d9162b61d8/
  https://www.data.gouv.fr/fr/datasets/cas-confirmes-dinfection-au-covid-19-par-region/
  https://www.data.gouv.fr/fr/datasets/donnees-relatives-a-lepidemie-du-covid-19/
  https://www.data.gouv.fr/fr/datasets/covid19-nouvelle-caledonie/
  https://www.santepubliquefrance.fr/maladies-et-traumatismes/maladies-et-infections-respiratoires/infection-a-coronavirus/articles/infection-au-nouveau-coronavirus-sars-cov-2-covid-19-france-et-monde
  https://geodes.santepubliquefrance.fr/#c=indicator
  https://www.data.gouv.fr/fr/datasets/taux-dactes-medicaux-sos-medecins-pour-suspicion-de-covid-19-pour-10-000-actes-medicaux/
  https://www.data.gouv.fr/fr/datasets/taux-dhospitalisation-parmi-les-passages-aux-urgences-pour-suspicion-de-covid-19/
  https://www.data.gouv.fr/fr/datasets/taux-de-passages-aux-urgences-pour-suspicion-de-covid-19/
  Tests
  https://www.data.gouv.fr/fr/datasets/donnees-relatives-aux-tests-de-depistage-de-covid-19-realises-en-laboratoire-de-ville/
  https://www.data.gouv.fr/fr/datasets/donnees-des-urgences-hospitalieres-et-de-sos-medecins-relatives-a-lepidemie-de-covid-19/

• Italy
  https://github.com/pcm-dpc/COVID-19/tree/master/dati-regioni
  CSV https://raw.githubusercontent.com/pcm-dpc/COVID-19/master/dati-regioni/dpc-covid19-ita-regioni.csv
  DOC https://github.com/pcm-dpc/COVID-19#dati-per-regione
  others:
  https://github.com/ondata/covid19italia#api

• US
  https://usafacts.org/visualizations/coronavirus-covid-19-spread-map/

• United Kingdom
  https://github.com/tomwhite/covid-19-uk-data/tree/master/data
  https://www.gov.uk/government/publications/covid-19-track-coronavirus-cases
  https://fingertips.phe.org.uk/documents/Historic%20COVID-19%20Dashboard%20Data.xlsx

• Germany
  RKI dashboard : https://experience.arcgis.com/experience/478220a4c454480e823b17327b2bf1d4/page/page_0/
  Underlying data :
  manually reported https://github.com/micgro42/COVID-19-DE/tree/master/time_series => seems the most practical robust and reliable
  more automated with extra sources https://github.com/jgehrcke/covid-19-germany-gae
  https://github.com/covid19-eu-zh/covid19-eu-data
  or also https://covid19-eu-data-cache.now.sh/covid-19-de.csv

• Spain
  Official dashboard with data https://covid19.isciii.es
  https://github.com/victorvicpal/COVID19_es/blob/master/data/final_data/dataCOVID19_es.csv
  https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/GPFFAQ ?

NOTAs announce:

El objetivo de los datos que se publican en esta web es saber el número de casos acumulados a la fecha y que por tanto no se puede deducir que la diferencia entre un día y el anterior es el número de casos nuevos ya que esos casos pueden haber sido recuperados de fechas anteriores. Cualquier inferencia que se haga sobre las diferencias de un día para otro deben hacerse con precaución y son únicamente la responsabilidad del autor.,,,,,,,,
Los datos de estas comunidades son datos de prevalencia (personas ingresadas a fecha de hoy). No reflejan el total de personas que han sido hospitalizadas o ingresadas en UCI  a lo largo del periodo de notificación(CL(UCIs)-GA(UCIS)-CM*-MD**),,,,,,,,
* Desde el día 11/04/2020 las cifras de hospitalizados de CM son casos acumulados. Previamente se refieren a personas ingresadas ese día.,,,,,,,,
* Desde el día 12/04/2020 las cifras de UCIs de CM son casos acumulados. Previamente se refieren a personas ingresadas ese día.,,,,,,,,
* Desde el día 26/04/2020 las cifras de Hospitalizados y UCIs de CM son casos acumulados. Previamente se refieren a personas ingresadas ese día. Se actualizará la serie cuando esté disponible,,,,,,,,
NOTA2:Se excluyen de la serie las notificaciones de personas  con anticuerpos positivos sin síntomas en el momento de realización de la prueba en los que no se puede establecer un momento de contagio ni si han padecido o no la enfermedad.,,,,,,,,      

There are also issues with confirmed cases values which are not reported anymore on many lines
This can be seen for instance here: https://boogheta.github.io/coronavirus-countries/#country=Spain&confirmed&recovered&hospitalized&intensive_care&deceased&log&multiples&places=Castilla%20La%20Mancha,Entire%20Spain,Galicia,Madrid&alignTo=deceased

Since I'm at it, it would be great if the graph colors could be more easily be distinguished. It's sometimes nearly impossible to tell at which graph you are looking at.

It appears countries have a fixed color. Maybe there could be an alternative, where a fixed palette of colors, beginning with the most different ones (straight blue, red, green, ..) which would then get sequentially assigned to a dataset only after it got selected. The downside is, that colors would change by the order in which you select sets. But for me, this would still lead to a better experience.