If you suffer from jittery GPS data and from Strava consequently overestimating your athletic abilities (especially for runs), run this software for a more realistic account.
The activity data is exported from your Strava account to a gpx file and re-uploaded after applying a smoothening algorithm.
Web interface
http://waaterbron.hopto.org/smoothify Follow the link to smoothen your data through the web interface. The generated information is not stored in a database. If you want to run the program on your own device follow the steps below.
Installation
pip install git+https://github.com/jonderwaater/smoothify You need an access token. You can generate it on http://waaterbron.hopto.org/smoothify. Store the token in ~/.smoothify as described above the generated token.
General usage
When your activity is uploaded to your account, run the software in a shell:
$ smoothenYour original activity will be replaced automatically. (Update 24.04.2017: changed permissions by Strava broke the functionality of this python program. Use the web interface and delete the activity on Strava before clicking the upload button on Smoothify to replace your activity.)
In the directory where you run smoothen the following files are created:
[Activity id].gpx - the gpx file of your original activity
[Activity id]_smooth.gpx - the gpx file with smoothened data
[Activity id].png - plot comparing the original and the smoothened data
Warning:
This code has only been tested on activities recorded with the Strava Android App.
Currently only position, time and elevation data are extracted, as well as the activity type, name and description.
That means: no heartrate data, kudos, comments, photos etc. This information from the Strava activity will be lost in the upload process unless you back it up by hand.
Options
-h , --help show this help message and exit
-a <id> , --activity-id <id> activity id; latest activity is taken by default.
-g , --no-graph do not create png file showing original and smoothened data.
-u , --no-upload do not upload smoothened data to Strava.
-d , --dry-run do not overwrite original activity when attempting to upload to Strava.
-c , --cleanup delete created files after uploading activity.
-s <algo> , --algo <algo> smoothening algorithm. running average (-s=1) is default.
On Android
The program will also run on Android on the Termux app. The installation method is identical to the one described above. Because of limited availability of packages it is required to run with option -g.
The scripts
-- extractgpx.py
Retrieves data from the latest Strava activity and writes it to a gpx file. As far as I know there is no other available method to obtain a gpx file from the command line.
-- smoothen.py
Smoothens the data in the gpx file. Currently it uses a simple algorithm that takes running average of 9 points, in steps of 1 point. More advanced algorithms may be implemented in the future, but it's already a significant improvement over untreated data.
-- upload.py
Uploads the the activity.
-- functions.py
Contains helper functions
Dependencies
stravalib, matplotlib, gpxpy and others
OTHER SOURCES:
For uploading using stravalib lessons were learned from https://github.com/dlenski/stravacli.
For plotting gpx files I looked at a blog post by Andy Kee.