nrjones8 / sf-ballot-measures

R Shiny app for visualizing historical ballot measures in SF - includes dataset of all past ballot measures (and their outcomes) in San Francisco elections

Historical Ballot Measures Data for San Francisco

The San Francisco Public Library (SFPL) maintains historical records of ballot propositions from San Francisco elections over the years. That database can be found on the SFPL website. However, it can be difficult to navigate or look at structured data about these elections and results. The Department of Elections maintains some structured data on past propositions, but the dataset only includes basic information (e.g. the prop letter, what % voted for it, and the title). It does not contain a more detailed description of the propositions, or links to further information.

That's where this repo comes in! The generate_ballot_measures_csv.py script scrapes the SFPL's website to build a structured CSV of past propositions - including their full description, and links to individual SFPL pages that describe the proposition in more detail. That CSV can be found in the data directory of this repo.

sf-ballot-measures's Issues

add heatmap, maybe on homepage

add google analytics

add basic search

data isn't (aren't, whatever) that big, can probably just do something like: https://www.gatsbyjs.org/docs/adding-search/#use-an-open-source-search-engine

and index the title + description of the prop. maybe allow additional keywords to be added manually too (e.g. searching "airbnb" should come up with Prop F from 2015)

deploy static site somewhere

https://www.gatsbyjs.org/tutorial/part-one/#deploying-a-gatsby-site

they suggest surge, worth trying? does it use a CDN or does it even really matter for this small of a thing?

just use that for now, could look into S3 later if really want to (https://www.gatsbyjs.org/docs/deploying-to-s3-cloudfront/)

add reminder to vote / links to SF elections website

add "by prop letter" summary page

/prop-letter/B or something

page displays summary stats of how prop As have performed over time (i.e. 70% have passed, maybe with links to individual elections).

Probably generate those summary stats in Python or elsewhere, spit out JSON from there, and just import those stats directly rather than calculating them in JS?