WeRateDogs_Twitter

This project was completed as part of Udacity's Data Analyst Nanodegree Program certification in January 2021.

Summary

Using Python libraries, I gathered data from a variety of sources and in a variety of formats, performed data wrangling on these datasets by assessing their quality and tidiness, then cleaned them. Finally, I performed Exploratory Data Analysis (EDA) on the clean and tidy dataset and presented my insights.

The dataset are the tweet archive of Twitter user @dog_rates, also known as WeRateDogs. WeRateDogs is a Twitter account that rates people's dogs with a humorous comment about the dog. These ratings almost always have a denominator of 10 and a numerator of greater than 10. Why? Because "they're good dogs Brent." WeRateDogs has over 8.9 million followers and has received international media coverage.

Data Wrangling Process

• Gathered data from different sources (Udacity WeRateDogs Twitter archive, retweet count and favorite count from Tweepy and tweet image predictions) and documented unclean issues first
• Assessed the data files for quality and addressed completeness, validity, accuracy and consistency of the available data
• Assessed the data for tidiness to make sure that structural or organizational issues are addressed
• Cleaned the data by fixing the quality and tidiness issues that were identified in the previous step, using Python and Pandas
• The assessment process started with visual assessment and was followed by programmatic assessment

Technologies Used

• Python, Numpy, Pandas, Matplotlib, Seaborn, JSON, Tweepy
• Jupyter Notebook, Regex

Insights from EDA

• The number of retweets has a strong positive correlation with the number of favorites a tweet receives. Both of these variables are also positively correlated with the rating that the dogs have received (although not as strong of a correlation)
• Based on the data available, pupper is the most popular dog stage, followed by doggo, puppo and floofer
• The highest rated dog breeds are Rottweiler, French_bulldog, Old_english_sheepdog, Golden_retriever, Bloodhound, Samoyed, Chihuahua, Irish_setter, Pembroke, Standard_poodle, tan_coonhound, Labrador_retriever, Pomeranian, Pomeranian, Lakeland_terrier, Eskimo_dog, Bedlington_terrier and Gordon_setter

Resources

Tweepy API Reference

RegexOne

Python Requests Library

Python DataFrame Constructor

Reading and Writing JSON to a File in Python

Using a code timer

Handling Exceptions

Try and Except in Python

How to Scrape More Information From Tweets on Twitter

Using conditional to generate new column in pandas dataframe