Dash Dashboard for Group Strava Activity

Credits

This project has been massively helped by the following people:

• Felix Flath (felix.flath@uniklinik-freiburg.de);
• Joachim Greiner (joachim.greiner@uniklinik-freiburg.de);
• Pavamkumar Videm (videmp@informatik.uni-freiburg.de);
• the Early Career Scientists from CRC1425.

Introduction

This project has been changed heavily for the event held. The following states the major changes:

• Instead of uploading to a Google Sheets, the program now displays the data on a webpage (hosted by PythonAnywhere);
• Instead of running the entire script every time to update the data and the table, a script has to be run once to create the main data table, and another every time the data table has to be updated;
• Instead of tracking kilometers, kilocalories are tracked;
• Script now based on requests library.

This project collates data from athletes' Strava account after the application has been authorised. The data is then organised by activities, which are then displayed in on a Dash table.

Setup

Folder Structure

First, clone project. The project folder should look like the following:

strava_club_challenge_automated
├── assets
│   ├── athlete_access_tokens
│   └── strava_app.yaml
├── dash_folder
│   ├── 0_frontpage.py
│   ├── 1_table.py
│   ├── 2_athlete_comparison.py
│   └── 3_activity_breakdown.py
├── data
├── resources
│   ├── athlete_access_tokens
│   └── strava_app.yaml
├── src
│   ├── authentication
│   │   ├── __init__.py
│   │   └── strava_athlete_authentication.py
│   ├── update_strava
│   │   ├── __init__.py
│   │   ├── get_activity_dataframe.py
│   │   ├── only_update_pickles_and_data_table.py
│   │   └── refresh_tokens.py
│   └── __init__.py
├── main.py
└── README.md

Assets

Contains the assets needed for the webpage, e.g.: logos.

Dash Folder

Contains the files needed for the Dash dashboard. These are used in main.py. The main pages are a front page, frontpage.py and table.py. Additional data analysis pages can be added.

Data

Contains a folder called athlete_pickles, where the individual athlete pickles are stored within. Additionally, pickle containing all the collated data, final_table_pickle.pkl, and an Excel table listing all the athletes, their teams, and their Strava IDs, athlete_database.xlsx, are also stored here.

Resources

Contains the folder athlete_access_tokens, where the athlete access tokens are stored. Additionally, a yaml file containing the Strava app's strava_client_id and strava_client_secret is also stored here.

Source

Contains the script(s) for authenticating individual athletes and to update or to create the initial table.

Python Anywhere Hosting

TODO

Athletes' Authorisation

For the application to have the permission to scrape data from an athlete's Strava profile, the athlete first has to authorise the application:

1. The athlete has to first click on the following link:

   https://www.strava.com/oauth/authorize?client_id=[CLIENT_ID]&response_type=code&redirect_uri=http://localhost/exchange_token&approval_prompt=force&scope=profile:read_all,activity:read_all

2. After approving the app, the athlete would be redirected to an empty page;
3. The link would have to be copied and pasted into the script in src.authentication.strava_athlete_authentication.py;
4. If the authorisation was successful, one would receive a status code 200 and the access token would be stored in json_path;

If authorisation was unsuccessful, check if the required fields were authorised.

Creation of Initial Data Table

TODO

Updating of Data Table

Run src.update_strava.only_update_pickles_and_data_table.py when table should be updated. This is done a number of times daily. Update and refresh app on PythonAnywhere.

Adding New Athletes

After the data table final_table_pickle.pkl has been created, new participants can be retroactively added:

1. Authorise athlete as per above steps;
2. Add athlete and athlete ID (and team) into athlete_database.xlsx;
3. Run src.update_strava.only_update_pickles_and_data_table.py.

Room for Improvement

The following items could be improved on (based on feel and user feedback):

• Improve look and formatting of table;
• Improve user authorisation process;
• Automation of updating process, perhaps to make it inline with Strava's recommendation of using webhooks.