This is a fork of the everylotbot library by Neil Freeman. It is a Twitter bot that tweets a Google Street View image of every single lot, whether vacant or not, in Tallahassee, FL. It is currently running at @everylottallahassee. This bot has no particular purpose. I just hope it brings you some value and illuminates your perception of spatial disparities -- read more about that here.
- Python 3.x
- A fresh Twitter account and a Twitter app, with registered keys
- A Google Streetview API token.
- A csv file with a row for every address you're interested in
- A place to run the bot (like a dedicated server or a GCP virtual machine)
Creating a Twitter account should be straightforward. To create a Twitter app, register at apps.twitter.com/. Once you have an app, you'll need to register your account with the app. Twitter has details.
Visit the Google Street View Image API page and get a key.
You'll need a CSV file of the lots. For @everylotintlh, I exported the "Download Certified Tax Roll Data" CSV from Leon County's Property Appraiser site and cleaned it up, as shown in the import.py file.
Here are the fields that were used by the bot:
- tweeted - to track amount of times the lot was tweeted
- address - for google street view lookups
- street - for composing tweets
You could use a virtual server hosted at a vendor like Amazon AWS or GCP. I used a GCP Virtual Machine and followed this tutorial to set it up.
Once you've created lots.csv -- a clean CSV of the lots (with a 'street', 'address' and 'tweeted' column) -- and configured your virtual environment or YAML file (this is what I used. example is in 'yaml.py' file) to house your Twitter/Google credentials in the setup.py file, you're ready to go.
Run: 'python3 bot.py' in the command line. It will post 1 lot every 30 minutes until it runs out of lots.