COVID-19 Data

COVID 19 cases data from Johns Hopkins, augmented and reformtted

jhu.edu-covid19-2.4.38. Modified 2020-07-08T04:59:37

Resources | Packages | Documentation| Contacts| References| Data Dictionary

Resources

• confirmed. Confirmed Non-US cases by date and country
• deaths. Non-US Death cases by date and country
• recovered. Non-US recoveries cases by date and country
• confirmed_us. Confirmed US cases by date and country
• deaths_us. US Death cases by date and country

Documentation

This dataset processed and augments the COVID-19 data provided by Johns Hopkins University Center for Systems Science and Engineering (JHU CSSE). The source data is checked into Github daily. and is collected from a variety of sources.

This dataset reformats the data into tidy format, with dates expressed as values instead of column headings, and adds several fields that are useful for analysis.

The ‘rate_t5d’ column is the growth date from 5 days before the observation to the observation. For example, for a row with a current observation of value x_5, and a past observation of x_0, the rate_t5d is calculated as e^((log(x_5)-log(x_0)) / 5)-1 . The result is that x_5 = x_0 * (1+rate_t5d)^5, and rate_t5d from the previous 5 days

Caveats

• China’s minimum cases in the dataset is 548, so it’s value for days sincle 100 cases is shifted by 6 days. It’s just a guess, but it looks good.
• Countries that haven’t reached 100 cases yet will have a days since 100 cases value that is always negative.

Documentation Links

• John Hopkins Project Homepage
• Documentation Page

Contacts

• Origin

  John Hopkins University Center for Systems Science and Engineering

• Wrangler

  Eric Busboom, Civic Knowledge

Data Dictionary

confirmed | deaths | recovered | confirmed_us | deaths_us

confirmed

Column Name	Data Type	Description
country	string	Country
province	string	Province, state, country or other region
location	string	Combination of county and province
date	datetime	Date of observation
confirmed	integer	Cumulative number of confirmed positives
date_10	datetime	Earliest date at which there were more than 10 cases
days_10	integer	Number of days since the earliest date of 10 cases
date_100	datetime	Earliest date at which there were more than 100 cases
days_100	integer	Number of days since the earliest date of 100 cases
rate_t5d	number	Growth rate, averaged over the following 5 days.
confirmed_log	number	Log of the number of confirmed positive cases

deaths

Column Name	Data Type	Description
country	string	Country
province	string	Province, state, country or other region
location	string	Combination of county and province
date	datetime	Date of observation
death	integer	Cumulative number of deaths
date_10	datetime	Earliest date at which there were more than 10 cases
days_10	integer	Number of days since the earliest date of 10 cases
date_100	string	Earliest date at which there were more than 100 cases
days_100	integer	Number of days since the earliest date of 100 cases
rate_t5d	number	Growth rate, averaged over the following 5 days.
death_log	number	Log of the number of deaths

recovered

Column Name	Data Type	Description
country	string	Country
province	string	Province, state, country or other region
location	string	Combination of county and province
date	datetime	Date of observation
recovered	integer	Cumulative number of recoveries
date_10	datetime	Earliest date at which there were more than 10 cases
days_10	integer	Number of days since the earliest date of 10 cases
date_100	datetime	Earliest date at which there were more than 100 cases
days_100	integer	Number of days since the earliest date of 100 cases
rate_t5d	number	Growth rate, averaged over the following 5 days.
recovered_log	number	Log of the number of recoveries

confirmed_us

Column Name	Data Type	Description
uid	integer
date	datetime
confirmed	integer
date_10	datetime
days_10	integer
date_100	string
days_100	integer
rate_t5d	number
confirmed_log	number
iso2	string
iso3	string
code3	integer
fips	integer
admin2	string
province	string
country	string
lat	number
long	number
location	string
population	integer

deaths_us

Column Name	Data Type	Description
uid	integer
date	datetime
death	integer
date_10	datetime
days_10	integer
date_100	string
days_100	integer
rate_t5d	number
death_log	number
iso2	string
iso3	string
code3	integer
fips	integer
admin2	string
province	string
country	string
lat	number
long	number
location	string
population	integer

References

Urls used in the creation of this data package.

• ts_base_url. Base URL for time series data
• confirmed_ts_source. Source for time series of confirmed cases, excluding US
• death_ts_source. Source for time series of deaths, excluding US
• recov_ts_source. Source for time series of recoveries, excluding US
• confirmed_ts_us_source. Source for time series of confirmed cases, US Only
• death_ts_us_source. Source for time series of deaths, US Only

Packages

• s3 s3://library.metatab.org/jhu.edu-covid19-2.4.38.csv
• csv http://library.metatab.org/jhu.edu-covid19-2.4.38.csv
• source https://github.com/metatab-packages/metatab-packages.git

Accessing Data in Vanilla Pandas

import pandas as pd


confirmed_df =  pd.read_csv('http://library.metatab.org/jhu.edu-covid19-2.4.38/data/confirmed.csv')
deaths_df =  pd.read_csv('http://library.metatab.org/jhu.edu-covid19-2.4.38/data/deaths.csv')
recovered_df =  pd.read_csv('http://library.metatab.org/jhu.edu-covid19-2.4.38/data/recovered.csv')
confirmed_us_df =  pd.read_csv('http://library.metatab.org/jhu.edu-covid19-2.4.38/data/confirmed_us.csv')
deaths_us_df =  pd.read_csv('http://library.metatab.org/jhu.edu-covid19-2.4.38/data/deaths_us.csv')

Accessing Package in Metapack

import metapack as mp
pkg = mp.open_package('http://library.metatab.org/jhu.edu-covid19-2.4.38.csv')

# Create Dataframes
confirmed_df = pkg.resource('confirmed').dataframe()
deaths_df = pkg.resource('deaths').dataframe()
recovered_df = pkg.resource('recovered').dataframe()
confirmed_us_df = pkg.resource('confirmed_us').dataframe()
deaths_us_df = pkg.resource('deaths_us').dataframe()