Skip to content
View samforwill's full-sized avatar
  • New York, NY
  • 18:57 (UTC -04:00)

Block or report samforwill

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
samforwill/README.md

Hi There! Welcome to my GitHub.

Take a look at my Portfolio Page. Explore interactive dashboards and step-by-step project walkthroughs. I also put together some smaller projects that don't have their own repos. Plus, a little about me.

About Me

I use data to support progressive campaigns and civic infrastructure. My work includes voter behavior modeling, precinct-level mapping, and demographic targeting using Python, SQL, and GIS.

Based in Queens, always chasing better bike lanes.

Developed for the newly redistricted Congressional Districts following the 2020 Census, this model leverages the Census Bureau’s 2021 American Community Survey data to predict Partisan Voter Index (PVI) scores with high accuracy for all US congressional districts. The accompanying Streamlit app visualizes key demographic influences on PVI scores and simulates electoral outcomes based on different turnout scenarios.

This project analyzes twitter messaging strategies from Democratic candidates who won difficult districts in the 2022 midterms. I used Non-Negative Matrix Factorization, NMF, and GloVe vector embeddings to uncover campaign strategies, such as Marie Gluesenkamp-Perez's focus on volunteer mobilization, and Chris Deluzio's strategy of tying his moderate opponenent to extreme figures. The insights from this analysis highlight tactical approaches that could influence election outcomes.

Employing demographic-based modeling, this project assesses the partisan leanings across state senate districts. Utilizing QGIS and GeoPandas, I mapped these leanings to visualize and predict electoral behaviors effectively.


Pinned Loading

  1. nlp-campaign-messaging nlp-campaign-messaging Public

    Analyzing the Twitter messaging strategies of two congressional candidates who won difficult districts in 2022.

    Jupyter Notebook 1

  2. predict-cd-pvi predict-cd-pvi Public

    Building a party support model based on 451 unique demographics of every Congressional District in the US

    Jupyter Notebook 1

  3. sldu-pvi sldu-pvi Public

    Modeling predicted PVI scores for state legistlative districts and mapping with QGIS

    Jupyter Notebook 1