This repo contains the data, code, and analysis of the 2024 U.S. presidential election forecast, designed by Jiwon Choi and Kevin Roe. Parquet file was not able to be used due to unresolved technical issues.
The repo is structured as:
data/00-simulated_datacontains the simulated data.data/01-raw_datacontains the raw data as obtained from fivethirtyeight.data/02-analysis_datacontains the cleaned dataset that was constructed.modelcontains fitted generalized linear models.othercontains sketches.papercontains the files used to generate the paper, including the Quarto document and reference bibliography file, as well as the PDF of the paper.scriptscontains the R scripts used to simulate, clean, and test data. Also, the folder contains R script for data modeling.
You can get data about polling outcomes from fivethirtyeight. To download the data, search for “Download the data”, then select Presidential general election polls (current cycle), then “Download”.
None of the components of this work involved the use of LLMs.