For Fun

I enjoy using tools like RShiny and Plotly to create interactive visualizations and applications that highlight aspects of my analysis and make them more enjoyable and informative to interact with. Stay tuned for more updates to this section of my site!

twittR

6/17

  • RShiny web app that analyzes a given Twitter user or hashtag based on the bag-of-words model.
  • Includes word plots of n-grams of length 1-4, word correlation plot, topic modeling and association using Latent Dirichlet Allocation (LDA), and tweet generation using Markov chains.
  • Read more and check out the source code here.

PokeSim R

8/17

  • RShiny web app that determines alikeness between Pokemon through a cosine similarity measurement generated from a numeric matrix of stats.
  • Created as part of an analysis of Pokemon data; view full analysis here.
  • Read more and check out the source code here.

For School

Pitchfork Review Analysis

12/17

  • Final project for Linguistics 409 - Introduction to Computational Linguistics.
  • Data was scraped from pitchfork.com in Python using BeautifulSoup, then EDA was performed in R using the tidyverse and language processing was done in Python with NLTK.
  • For a more extensive analysis using R, view my blog post Pitchfork Reviews: 1999-2017 here.

Determining What Factors Most Impact the Health of a Newborn Child

12/16

  • Final project for Public Health 490ST - Telling Stories with Data.
  • Our group’s introduction and conclusion, as well as the results of my analysis, are included.
  • I focus on fitting a logistic regression model to predict whether a birth will be unhealthy or not based a variety of demographic and biological predictors, specifically exploring the mother’s race and the number of children born at once.