rjp / stravathings

A random bunch of tools for www.strava.com (cycle-mapping thing)

A reasonably random collection of scripts relating to Strava

Notes

Requirements

I'm using Python 2.7 (2.6 will might work also)

• unum is used to prevent silly unit-conversion related errors
• redis is used for caching, although as discussed in the caching section, this requirement could be removed quite easily.

There is a requirements.txt file containing the specific versions (so you can pip install -r requirements.txt)

Athlete ID

Most scripts default to my Strava athlete-id

You probably want to modify the myid function in common.py to your ID (found by going to your profile page, and looking at the number in the URL, e.g given http://app.strava.com/athletes/316985 the ID is 316985

Most scripts should take an argument to change this (usually -a 1234), although changing the myid function saves from always having to type this

Caching

Since a lot of these scripts make a lot of requests to the Strava API (usually one or two per ride, on each run), there is some basic caching in place.

The caching is currently based on redis, because it was easy to implement.. For it to work, you need to be running a redis server, and have the redis Python module installed. On OS X with homebrew installed:

$ brew install redis
$ pip install redis

All URL-retrival is done via geturl in common.py, so implementing a file-based cache would be easy..

Scripts

allrides_map

For a given athlete, creates a self-contained HTML file, with all routes cycled overlaid on a single map.

$ python strava_allrides_map.py

By default this grabs the last 4 pages of rides (which means 200 rides). You can change this limit with -l 10 which will download 500 etc.

strava_sortomatic.py

Prints your last 50 rides, ordered by a specific attribute like averageSpeed

$ python strava_sortomatic.py averageWatts | tail -11
# Crafers
18.36 km (0.77 hours moving)
24.00 km/h avg (57.08 km/h max)
503.27m climbed
199w average power

# Last cycle before back to Scotland for a while
62.74 km (2.76 hours moving)
22.73 km/h avg (50.66 km/h max)
305.00m climbed
227w average power

strava_multisegment.py

Based on your last 50 rides, shows all segments you've done more than once, along with your times for each.

The most recent time should be on the right, although I wouldn't rely on the ordering...

strava_dist_vs_elevation.py

Tries to sort the last 50 rides based on distance and amount of climbing.

I'm sure there's better ways to order them, but elevationgain squared, multiplied by distance seemed to work quite well

For example:

$ python strava_dist_vs_elevation.py 
164.4 km dist, 2433.48 up, Murray Bridge and back
93.2 km dist, 1360.78 up, Lofty, Corkscrew, Gorge and beach w/Scott
56.6 km dist, 1716.41 up, Mt Lofty figure of... 9
135.7 km dist, 1092.08 up, Around The Forth (NCN route 76'ish)

So the long ride with lots of climbing was first. The shorter ride with a relatively large amount of climibing is third, and a longer with not much climbing was 4th.

Strava challenge-specific scripts

There's a couple of scripts to track my progress on a few of Strava's challenges, which are pretty useless now, but could be modified for future things

Misc'ery

parse_gpx.py

This parses a GPX file and prints some basic, crudely formatted info like distance and so on.

None of this has been used, as the Strava API provides all the necessary numbers, so calculating speed from GPS data-points hasn't been necessary.

This uses Untangle for XML parsing, along with python-dateutil to parse the timestamps. Also unum to prevent stupid errors when converting the metres/sec to more readable units and such..

streetview.py

An incomplete experiment to take a GPS track and create a video of the route using Google Streetview images.

You need frames quite often along the route to maintain any consistency, and there is a limit of 1000 requests/hour for the Streetview API. Doable, just need to tweak the interval at which is grabs images, and there might be something wonky with the heading-calculation

ridetrim/ridetrim.html

A crude and incomplete web/Javascript application, for removing a bad section in the middle of a GPX file (say if you don't pause the GPS recording while in a car, or go in a building and the GPS position jumps around due to the weak signal)

You paste the contents of a GPX file into the first textarea, click "Load" button.

It shows the route on a map. You adjust the sliders to select a region of the track.

Then click "trim", and in the right-hand textarea, the new trimmed GPX file contents is displayed.

This means you could export a GPX from Strava, remove a noisy section of the ride, and create a clean GPX file. Then delete the original ride and reupload

Note that the tool will currently drop additional data like heart-rate, power and cadence.

Ideally this functionality would be incorporated into Strava's crop tool someday.