Usage

Local setup

$ virtualenv3 env
$ env/bin/pip install -e .
$ env/bin/debug-server.py /tmp/routemaster-db

Testing

$ env/bin/python -m unittest

Deploying to routemaster.lumeh.org

Install Fabric on your system and run

$ fab deploy

API

See routemaster/db.py for the definitions of the models. Accounts and journeys are identified by random UUIDs, with the exception of the test Account ("Hermann Dörkschneider"), whose id is test.

Sessions

The requests will contain an accountId header which the server will use to determine how certain requests are handled.

Fetching data

• GET /account/<uuid>

• GET /account/<uuid>/recent

  Returns a list of the account's recent public Journeys.

• GET /journey/<uuid>

  Can return 403 if the Journey is not public.

• GET /route/<int:id>/top

  Returns a list of the top-scoring journeys for the route.

• GET /route/<int:id>/recent

  Returns a list of the most recent journeys for the route.

Storing data

• POST /journey – store a recorded journey

  Required data:

  {
    "id": ...,
    "visibility": ...,
    "startTimeUtc": ...,
    "stopTimeUtc": ...,
    "startPlaceId": ...,
    "stopPlaceId": ...,
    "waypoints": [
      {
        "timeUtc": ...,
        "accuracyM": ...,
        "latitude": ...,
        "longitude": ...,
        "heightM": ...
      },
      ...
    ]
  }
  

• POST /account – create a new account ?

Scoring

The efficiency score is an integer from 0 to 100. We first calculate the straight-line distance between the starting and stopping locations. Then we calculate the distance the user walked by summing the distances between adjacent recorded waypoints. The efficiency score is then

min(0, ceil(200 * (straight-line distance)^2 / (walked distance)^2 - 100))

The purpose of the min(0, (...) - 100) is to make it easier to get low scores. Previously the formula was just (100 * ...), which meant you'd have to walk an infinate distance in order to get a score of 0.