Gym Forecasting (WIP)

Forecasting how busy my local gym will be using Prophet

Install

pip install -e .

Fetching gym occupancy data

Fetch data from TheGymGroup's API through an authenticated session using membership credentials.

from src.data.gym import start_gym_session, get_gym_ids, get_gym_data, save_gym_data

# username should be your TheGym membership email
# password should be your TheGym membership PIN (8-digits as a string, e.g.: "12345678")
sess = start_gym_session(username="your_email@provider.com", password="your_PIN")

A comprehensive list of 209 TheGym branchs ids (required for fetching data from TheGym's API) can be loaded.

# retrieve dictionary of branch:id pairs for 209 TheGym branches
gyms = get_gym_ids()
for branch, identifier in gyms.items():
    print(branch, identifier)

> Altrincham 057d7a7c-a61a-423f-b21f-54adedc1cd0d
> Ashford 5cc9b50b-c049-49d2-a929-b9e9ec34f9ba
> ...205 others...
> Worthing 23becc48-441d-454e-b995-69a88b4853f4
> York Foss Island 942ccf8b-b855-4e9a-928a-99a2513a329a

Finally, retrieve and save data for the gym branch of interest

data = get_gym_data(sess, gyms["London Angel"])  # returns data as JSON
save_gym_data(data)  # appends data to `./data/interim/gym.csv`

Visualising gym occupancy data

See the README notebook for src.visualisation.plotting examples.

Forecasting gym busyness (WIP)

Prophet - Simple example with weekdays and weekends/bank holidays.