Bing COVID-19 Dataset

Access daily updated COVID-19 data including confirmed, fatal, and recovered cases from all regions worldwide.

Bing COVID-19 Dataset

Bing COVID-19 data includes confirmed, fatal, and recovered cases from all regions, updated daily. The Bing COVID-19 Tracker reflects this data.

Data Sources

Bing collects data from multiple trusted, reliable sources, including:

• BNO News
• Centers for Disease Control and Prevention (CDC)
• National/regional and state public health departments
• Wikipedia
• World Health Organization (WHO)
• 24/7 Wall St.

Dataset Information

Available Formats

Modified Bing COVID-19 datasets are available in multiple formats:

• CSV (560.3 MB)
• JSON (1515.6 MB)
• JSON-Lines (1506.2 MB)
• Parquet (55.4 MB)

All modified datasets have ISO 3166 subdivision codes and load times added. They use lowercase column names with underscore separators.

CSV-format raw data is also available.

Data Volume

• Total Records: 4,766,737+ rows (as of March 5, 2023)
• Update Frequency: Daily
• Historical Data: Earlier versions available on GitHub

Schema

Column Name	Data Type	Description	Example Values
id	int	Unique identifier	742546, 69019298
updated	date	The as-of date for the record	2021-04-23
confirmed	int	Confirmed case count for the region	1, 2, 1000
confirmed_change	int	Change in confirmed cases from previous day	1, 2, 50
deaths	int	Death case count for the region	1, 2, 100
deaths_change	smallint	Change in deaths from previous day	1, 2, 5
recovered	int	Recovered count for the region	1, 2, 500
recovered_change	int	Change in recovered cases from previous day	1, 2, 25
country_region	string	Country/region	United States, India
admin_region_1	string	Region within country_region	Texas, Georgia
admin_region_2	string	Region within admin_region_1	Washington County
iso2	string	2-letter country code identifier	US, IN
iso3	string	3-letter country code identifier	USA, IND
iso_subdivision	string	Two-part ISO subdivision code	US-TX, US-GA
latitude	double	Latitude of the region centroid	42.28708, 19.59852
longitude	double	Longitude of the region centroid	-2.5396, -155.5186
load_time	timestamp	Date and time file was loaded from source	2021-04-26 00:06:34

Preview Data

id	updated	confirmed	deaths	country_region	admin_region_1	confirmed_change	deaths_change
338995	2020-01-21	262	0	Worldwide	null	-	-
338996	2020-01-22	313	0	Worldwide	null	51	0
338997	2020-01-23	578	0	Worldwide	null	265	0
338998	2020-01-24	841	0	Worldwide	null	263	0
338999	2020-01-25	1320	0	Worldwide	null	479	0
339000	2020-01-26	2014	0	Worldwide	null	694	0
339001	2020-01-27	2798	0	Worldwide	null	784	0
339002	2020-01-28	4593	0	Worldwide	null	1795	0

Data Access

Python (Pandas Direct Access)

Download and analyze the dataset using pandas:

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Read directly from Azure Blob Storage
df = pd.read_parquet(
    "https://pandemicdatalake.blob.core.windows.net/public/curated/covid-19/bing_covid-19_data/latest/bing_covid-19_data.parquet"
)

print(df.head(10))
print(df.dtypes)

Analyze Worldwide Data

# Filter worldwide data
df_worldwide = df[df['country_region'] == 'Worldwide']

# Create pivot table
df_worldwide_pivot = df_worldwide.pivot_table(
    df_worldwide, 
    index=['country_region', 'updated']
)

print(df_worldwide_pivot)

Visualize Trends

# Plot confirmed cases over time
df_worldwide.plot(
    kind='line',
    x='updated',
    y='confirmed',
    grid=True,
    title='Worldwide Confirmed COVID-19 Cases'
)

# Plot deaths over time
df_worldwide.plot(
    kind='line',
    x='updated',
    y='deaths',
    grid=True,
    title='Worldwide COVID-19 Deaths'
)

# Plot daily change in confirmed cases
df_worldwide.plot(
    kind='line',
    x='updated',
    y='confirmed_change',
    grid=True,
    title='Daily Change in Confirmed Cases'
)

plt.show()

Azure Databricks (PySpark)

Access the dataset in Azure Databricks:

# Azure storage access configuration
blob_account_name = "pandemicdatalake"
blob_container_name = "public"
blob_relative_path = "curated/covid-19/bing_covid-19_data/latest/bing_covid-19_data.parquet"

# Configure SPARK to read from Blob remotely
wasbs_path = f"wasbs://{blob_container_name}@{blob_account_name}.blob.core.windows.net/{blob_relative_path}"
print(f'Remote blob path: {wasbs_path}')

# Read parquet file
df = spark.read.parquet(wasbs_path)
df.createOrReplaceTempView('covid_data')

# Display top 10 rows
print('Displaying top 10 rows:')
display(spark.sql('SELECT * FROM covid_data LIMIT 10'))

Azure Synapse Analytics

# Azure storage access configuration
blob_account_name = "pandemicdatalake"
blob_container_name = "public"
blob_relative_path = "curated/covid-19/bing_covid-19_data/latest/bing_covid-19_data.parquet"

# Configure SPARK to read from Blob remotely
wasbs_path = f"wasbs://{blob_container_name}@{blob_account_name}.blob.core.windows.net/{blob_relative_path}"
print(f'Remote blob path: {wasbs_path}')

# Read parquet file
df = spark.read.parquet(wasbs_path)
df.createOrReplaceTempView('source')

# Display top 10 rows
display(spark.sql('SELECT * FROM source LIMIT 10'))

Analysis Examples

Query Specific Country Data

import pandas as pd

df = pd.read_parquet(
    "https://pandemicdatalake.blob.core.windows.net/public/curated/covid-19/bing_covid-19_data/latest/bing_covid-19_data.parquet"
)

# Filter for a specific country
us_data = df[df['country_region'] == 'United States']

# Get latest data by state
latest_date = us_data['updated'].max()
latest_by_state = us_data[
    us_data['updated'] == latest_date
].groupby('admin_region_1').agg({
    'confirmed': 'sum',
    'deaths': 'sum',
    'recovered': 'sum'
}).sort_values('confirmed', ascending=False)

print(f"\nTop 10 US States by Confirmed Cases (as of {latest_date}):")
print(latest_by_state.head(10))

Calculate Mortality Rate

# Calculate mortality rate by country
latest_global = df[df['updated'] == df['updated'].max()].copy()
latest_global['mortality_rate'] = (
    latest_global['deaths'] / latest_global['confirmed'] * 100
).round(2)

top_countries = latest_global[
    latest_global['confirmed'] > 1000000
].sort_values('mortality_rate', ascending=False)

print("\nCountries with >1M cases - Mortality Rates:")
print(top_countries[['country_region', 'confirmed', 'deaths', 'mortality_rate']].head(10))

Time Series Analysis

# Analyze 7-day moving average of new cases
us_nationwide = df[
    (df['country_region'] == 'United States') & 
    (df['admin_region_1'].isna())
].sort_values('updated')

us_nationwide['7day_avg_cases'] = us_nationwide['confirmed_change'].rolling(window=7).mean()
us_nationwide['7day_avg_deaths'] = us_nationwide['deaths_change'].rolling(window=7).mean()

# Plot 7-day moving averages
import matplotlib.pyplot as plt

fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(12, 8))

ax1.plot(us_nationwide['updated'], us_nationwide['7day_avg_cases'])
ax1.set_title('US COVID-19: 7-Day Moving Average of New Cases')
ax1.set_ylabel('Cases')
ax1.grid(True)

ax2.plot(us_nationwide['updated'], us_nationwide['7day_avg_deaths'])
ax2.set_title('US COVID-19: 7-Day Moving Average of Deaths')
ax2.set_ylabel('Deaths')
ax2.set_xlabel('Date')
ax2.grid(True)

plt.tight_layout()
plt.show()

License and Attribution

The data is available strictly for educational and academic purposes under these terms and conditions.

Valid Use Cases:

• Academic institutions
• Government agencies
• Medical research

Attribution Requirement:

Data used or cited in publications should include an attribution to “Bing COVID-19 Tracker” with a link to www.bing.com/covid.

This dataset is for educational and research purposes only. Always verify critical health information with official sources.

Related Datasets

COVID-19 Data Lake

Access additional COVID-19 datasets covering testing, hospital capacity, and mobility

ECDC COVID Cases

European Centre for Disease Prevention and Control COVID-19 data

COVID Tracking Project

US state-level testing and outcome data

Oxford Government Response

Government policy responses to COVID-19

Next Steps

Create ML Dataset

Learn how to create Azure ML datasets from this data

Browse Catalog

Explore other available datasets

Public Holidays

View public holidays dataset

Genomics Data

Explore genomics datasets

Contact

For questions or feedback about this or other datasets in the COVID-19 Data Lake, contact [e-mail].