marcmaceira/covid-19-mobility-tracker

Google Mobility Reports, reverse-engineered into a JSON / CSV API

Covid-19 Mobility Tracker

Note

Note: Google recently started publishing raw CSV data for this, and following it is recommended for all future updates.

This project would still continue update sources, but it might not be as realtime.

Google Mobility Reports show aggregate activity in each country, and how it changes in response to policies aimed at combating COVID-19. However, it is only published as a PDF and the data isn't available in a machine-readable format that could enable more richer analysis.

This is an effort to reverse-engineer the PDFs into vectors and ultimately into time-series data available as a JSON Rest API.

Usage

Country-level data

Country level JSON is available for use at –

https://pastelsky.github.io/covid-19-mobility-tracker/output/<ISO-COUNTRY-CODE>/mobility.json

For eg India: https://pastelsky.github.io/covid-19-mobility-tracker/output/IN/mobility.json

US State-level data

US state level JSON is available at –

https://pastelsky.github.io/covid-19-mobility-tracker/output/US/<US-STATE-CODE>/mobility.json

For eg: New York: https://pastelsky.github.io/covid-19-mobility-tracker/output/US/NY/mobility.json

Accessing data as CSV

If you prefer to use CSVs instead, you can go to -

https://pastelsky.github.io/covid-19-mobility-tracker/output/<ISO-COUNTRY-CODE>/mobility-<social-place>.csv

Where social place is one of parks | residential | retail-and-recreation | transit-stations | workplaces | grocery-and-pharmacy

For eg. parks data for India in CSV - https://pastelsky.github.io/covid-19-mobility-tracker/output/IN/mobility-parks.csv

For US states, just add the state code after the country code.

Data structure

Each data element consists of date and value (which represents mobility change in percentage)

Limitations

• Expected error range is ±2%, though errors in activity percentanges should be rare given the methodolgy used to extract data.
• As of now, this only indexes country-level data and US state level data. State / Province level breakup for other countries is yet to be added.
• Data sources will be updated once / day and is incumbent upon google publishing updated reports
• It assumes that the same scale (-80% to +80%) is used in all PDFs and they are structured similarly.

Contributing

1. Run yarn install to install dependencies.

2. Install Inkscape 1.1-dev for your OS, and make sure the inkscape is available as in your path as a command line utility. You can test this using inkscape --version

3. Run yarn build to begin downloading reports form google, and parsing it into the ouput folder.

Credits

All data made available for use is by taken from Google Mobility Reports. This project does not claim any ownership over this data, and is not reponsible for guaranteeing accuracy of this data.