This project was completed as part of Udacity's Data Analyst Nanodegree.

WeRateDogs is a Twitter account that posts and rates pictures of dogs. These ratings often are not serious and have numerators that are greater than the denominators. In this analysis, I mostly focus on wrangling WeRateDogs's Twitter archive through August 1, 2017. Most of the necessary Twitter data has been provided by Udacity and includes information on each post, as well as details on each dog such as the name, rating, and stage (whether the dog is a doggo, floofer, pupper, or puppo).

In addition, Udacity ran the images on WeRateDogs's account through a neural network to generate three predictions for each image. For each prediction, there is also data on the confidence and whether the prediction is a type of dog breed.

In the data wrangling process, I only focus on original tweets by WeRateDogs that have ratings and images. At the end, I also provide a brief analysis using the cleaned data. In the analysis, I aim to answer the following questions:

• What is WeRateDogs's posting trend by month?
• What is the monthly trend of interactions with WeRateDogs's posts?
• What are the most popular dog breeds based on number of posts, interactions by Twitter users, and ratings?
• Is there any correlation between WeRateDogs's ratings and the interactions by Twitter users?

• Jupyter Notebook
• Pandas
• Numpy
• Requests
• Tweepy
• json
• Matplotlib
• Scipy
• Twitter API

To get the Jupyter Notebook running, execute the following in the command line and select wrangle_act.ipynb from the Jupyter Notebook dashboard. The conda environment setup is optional; I have provided the base environment in base.yaml.

$ git clone https://github.com/evanchen13/weratedogs.git
$ cd weratedogs
$ conda env create -f base.yaml
$ jupyter notebook

image-predictions.tsv is provided in this repository, but can be downloaded programmatically as shown in the code. Similarly, tweet_json.txt is provided in this repository, but can be downloaded programmatically using the Twitter API. This requires a developer account. To apply for access, visit this link and click the "Apply for a developer account" button. Then, be sure to edit the following lines in cell 5:

consumer_key = 'CONSUMER_KEY'
consumer_secret = 'CONSUMER_SECRET'
access_token = 'ACCESS_TOKEN'
access_secret = 'ACCESS_SECRET'