My best projects and work.
clubSandwich estimatr Compatibility Demo
Short demo page for a lightning talk regarding some of my contributions to clubSandwich, which allow some objects from estimatr to be passed to clubSandwich functions.
My Contributions: This is work from my internship with OSPO, so this page and much of the code being demonstrated was written by me.
US Migration Flows Visualization Project
View HTML | View PDF | Download HTML | Download PDF
This report explores US migration data from the years 2016 through 2020. All versions also include a link to a hosted shiny app.
My Contributions: Wrote much of the analysis, published our shiny app, cleaned up formatting, and proofread both code and writing. Additionally, conducted early literature review to assist groupmates in chosing the best packages for our visualizations.
Note: This report looks much better in HTML.
Exploration of Simulated Student Habit Data
View HTML | View PDF | Download HTML | Download PDF
This project is a brief exploration of a realistic simulated dataset of student habits, and it includes one of my favorite visualizations I have made.
My Contributions: This was a solo project, so all work is my own.
Note: This project looks much better in HTML.
Sentiment Analysis and Topic Modeling of Tweets
This research paper uses sentiment analysis and topic modeling to investigate tweets surrounding the 2020 presidential election.
My Contributions: Much of sections 1 through 3 were written or heavily edited by me, with 2 and 3 notably containing concise literature review. Wrote the code used to create most the figures in section 5, Results, and wrote the respective analysis for most figures. Wrote the entirety of section 6, Considerations and Future Directions, and most of section 7, Conclusion. Proofread all code and writing. Formatted document using LaTeX.
Deep Learning for Classifying Images of Egyptian Monuments
This report details explorations in image recognition through deep learning.
My Contributions: Much of the writing in sections 4-8, managing our GitHub repo, wrote our hybrid CNN-transformer model, wrote code to train and test all 12 variations of both models, and generated figures.
NYC Crime Analysis
View HTML | View PDF | Download HTML | Download PDF
This report provides visualizations and details statistical analysis done with New York City crime data.
My Contributions: Wrote all code used for data cleaning and preprocessing, and most of the code used to generate figures. Handled all formatting. Wrote and proofread parts of Introduction and Discussion, wrote most of Analysis, and wrote all of Background.
Note: This report looks much better in HTML.