Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on May 5, 2013
  1. Tests for the RK authentication flow

    Uses a mocked RK object.
  2. Start adding webapp tests

  3. Port to Flask-SQLAlchemy

    This resolves a lot of issues around setting up the test DB and handling
    multithreading in a much more sensible way.
    While we're solving threading/concurrency issues, don't use RK as a
Commits on May 4, 2013
  1. Make py.test use a different database (specifically cycleroutertest)

    Add a script which makes it easier to create the database.
  2. Move all database session management into a separate module

    This module must be imported and the Session initialised before you
    import the ORM, otherwise the ORM will not be able to interface with the
    The purpose of separating the session is that it allows us to use a
    separate database for the unit tests.
Commits on May 3, 2013
  1. Also add a created attribute to User

    This will allow us to track growth of the database.
Commits on Apr 28, 2013
  1. Test backrefs

  2. Test delete cascading

  3. Handle reimporting of tracks

  4. Relate Tracks to Users

  5. Add method User.to_rk()

  6. Test User.from_rk

  7. Mark lots of tests as skipped if they required plotting or the net

    Lots of skipped tests won't pass at the moment, that needs to be
    resolved, but I'm not sure how many of the tests are useful. Also I need
    to add some nice test data to the repo and use get_test_resource()
Commits on Apr 27, 2013
  1. orm: use raw JSON, not the numpy Track class

    Having thought about it, the ORM based Track class will probably replace
    the numpy version, especially if we want to generate an on-going
    updating map for a given part of the world.
  2. Experimental GeoAlchemy based ORM for storing Tracks

    Work in progress. Can import a Track and return it's length if you know
    the project to use.
Commits on Apr 6, 2013
  1. strm: extract_track method

    Method to extract the elevations from STRM using a Track object. Test to
    compare the calculated and observed elevations by direct comparison and
    calculating RMS.
  2. srtm: implement extract_point method

    Extracts the elevation of a single point from the SRTM data. Includes a
    test to compare the value against Google's elevation API.
Commits on Feb 24, 2013
  1. Plot outputs from STRM grid

    Looks like I'm not considering masked values.
  2. Load a GA SRTM DEM ESRI grid into a numpy array

    Uses GDAL to load the grid file into a numpy array. Calculates the
    lat/lons for the data coverage.
    Includes a skeleton context manager for loading GDAL files using 'with'.
    Doesn't yet do any error handling etc, or even close the dataset (the
    API seems to be lacking a close?).
    Requires a SRTM DEM (digital elevation model) data grid from Geoscience
    Australia. This data is Creative Commons, and can be acquired from
Commits on Feb 10, 2013
Commits on Jan 31, 2013
  1. Add directional binning

  2. Increase grid resolution

Commits on Jan 3, 2013
  1. Fix binning for x/y correctness

    Grid now matches what I would expect, tracing out my usual cycling
    routes through Melbourne.
  2. Consider all tracks, even those with manual entry

    Simply editing a track to say add a route, causes all the entries to be
    marked as 'manual' it seems. Completely manual tracks will have an
    evenly spread speed, so as a result their speed anomoly will be 0, and
    thus vanish from the binning.
  3. Rewrite bearing and distance calculations to use UTM zones

    This technique produces reliable bearings although does require you to
    specify the UTM zone.
Commits on Dec 26, 2012
  1. Add some annotations

  2. Prototype binning

    Doesn't yet deal with the idea that anomolies will be different in
    different directions
  3. Separate out GPX specific code from Track processing code

    Downloaded data is in JSON format
Something went wrong with that request. Please try again.