eecs398-python-twitter

Starter code used for live demo of integrating the Twitter API and Python for EECS 398 Winter 2018 at University of Michigan, Ann Arbor

Requirements

tweepy

$ pip3 install tweepy

• created specifically to pull data from Twitter

• no need to index into large dictionaries or format request URLs

matplotlib

pip3 install matplotlib

• good for plotting data

• "not the best out there, but easy to use" - Amrit

What's included?

bot.py contains starter code for the Python program

secret.py will hold your API keys and tokens

• you want to keep your API keys secret, which is why they are placed in a separate file from your program.

• if you choose to put your project on GitHub, DO NOT commit secret.py

How to Generate Twitter API Keys and Tokens

Go to https://apps.twitter.com

NOTE: You will need a Twitter account from this point on.

• sign into Twitter, then click "Create New App" in the upper right

• fill out the name, description, and website for your applicaiton. Check the "Developer Agreement" box and finish creating your app!

• under the name of your application, find and click the "Keys and Access Tokens" tab

Consumer Keys

• located under Applicaiton Settings

• copy and paste these keys into secret.py

Access Tokens

• scroll down and click "Authorize my Application"

• located under "Your Access Token"

• copy and paste these keys into secret.py

TODO: import libraries

• import tweepy

• import matplotlib

• from secret import *

TODO: create a new Twitter session

• check out the tweepy documentation on getting started

• use keys/tokens from secret.py to create the session

TODO: user input

• ask the user (you) what hashtag to search

TODO: search for hashtag

• create new Tweet objects for each result

• we only need the time each tweet was posted, but other data may be useful for debugging or other data analysis

TODO: graph timestamps from tweets in a histogram

• check out the matplotlib documentation on histograms and subplots

• this is just one analysis you could do on Twitter data

• try something on your own!