Skip to content
Some tools to read Mobi's public API
Jupyter Notebook Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
notebooks
LICENSE
README.md
geomobi.py
helpers.py
mobi.py
mobi_system_data.py
plots.py
vanweather.py

README.md

Mobi API query and analysis

Some tools for grabbing live data from the Mobi API and storing it in Pandas Dataframes.

Usage

Query API

python3 mobi.py --query /path/to/workingdirectory/

The first time this is run it will create a CSV file "daily_mobi_dataframe.csv" with a single row corresponding to the number of bikes currently at each station. Subsequent calls will add rows to the dataframe. I don't know how often Mobi updates the API, but I run this as a cron job every minute.

Update analysis dataframes

Continuously updating the dataframe via the above script quickly becomes inneficient, and loading the dataframe every minute will kill your computer's performance after a few days. So once a day I run a second script to break down the raw data to more manageable datasets.

python3 mobi.py --update /path/to/workingdirectory

This will make 6 new CSV files -- one each for bikes taken from, returned to, or both from each station, then grouped by either hours or days.

The minute-by-minute dataframe is backed up and renamed with the current date and time. Keep or remove at your discretion.

Secondary options

python3 mobi.py --stations /path/to/workingdirectory/

Create/update a CSV file with the number of available bikes at each station, or -1 for out-of-service stations. I run this at 4am daily (when there are close to zero ongoing trip) to get a snapshot of the total number of bikes and stations

python3 mobi.py --status /path/to/workingdirectory/

Prints the total number of bikes and active stations based on the last row in the csv file created with the --stations command.

Analysis

I will include Jupyter notebooks with analysis of this data as I complete them

Full dataset

If you would like access to the full dataset, please contact me directly: mike@mikejarrett.ca

You can’t perform that action at this time.