No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

Google maps travel time experiment

This project aims to use the travel time data provided by Google through it's Distance matrix API to perform experiments. The results can be found here

There is a location database and a results database, the locations are latatide and longitude pairs taken from GeoNames. The results come from taking random pairs or lat/long in the same country and calculating the distance and time as given by Google's maps API.

There is a limit of 2500 downloads per day from any given IP or API key, so the results are saved in this directory so anyone can contribute. To do so simply clone this repo, follow the installation instructions, then pick a country and start getting data. This is a Monte Carlo like study so the more data the better.


Install dependencies, if you have pip installed then

$ sudo pip install simplejson

$ sudo pip install pandas

Now we need to download the location database

$ cd Location_database


$ make

This will download the zip file and extract the data. Once done your all set simply cd ../ and do

$ ./basic_tools --help

to see the options

Basic usage:

  • Get a list of all countries available in the location database, this prints a list of the country codes, e.g GB=Great Britian

    $ ./basic_tools -g

  • Get a list of all the downloaded data so far

    $ ./basic_tools -l

  • Collect data, for example:

    $ ./basic_tools -r -N 1000 -c US

    The -r flag initiates the download data script, the -N requests 1000 attempts to collect data, and finally the country is given to -c by it's country code.

  • Plot the distance vs time graphs with the flag -p, then the countries you want to include

    $ ./basic_tools -p -c GB MX

  • Speed distrubution graphs can similarly be obtained from

    $ ./basic_tools -v -c US MX


  • python
  • simplejson
  • pandas
  • numpy
  • matplotlib