SF Fit Finder
An interactive map of San Francisco that overlays noise, weather, transit, parks, trees, and crime, then ranks neighborhoods against whatever you want more or less of.
I started this trying to check noise around SF, and ended up wondering which block was actually the best spot to live based on things like weather, walkability, transit, and crime. So now the map overlays all of that.
Live at sfdatamap.atomsclark.com.
You pick what matters at the top of the page and the neighborhood rankings shift. Open the rankings panel to tweak weights, or to pick warmer vs. cooler for the personal stuff like temperature and wind. There’s a small top-five strip that stays visible in the corner so you can flip a layer on and watch the order shuffle in real time.
The data comes from wherever it’s free. 311 noise complaints and SFPD incidents from DataSF, road density from OSM via Overpass, current temperature and air quality from Open-Meteo. No backend; everything fetches straight from the browser, with the heavier 311 buckets cached in IndexedDB so repeat visits are instant. State serializes into the URL, so any view I land on is a shareable link.
Vibecoded with Claude Code over a long weekend. Vite, React, TypeScript, MapLibre GL JS for the map, Dexie on top of IndexedDB for caching.
There’s some prior art on this. SF Planning aggregated a year of 311 noise complaints into a single map at one point, but it’s a static PDF. The SF Chronicle wrote up the Tenderloin’s share of those complaints using the same source. Both are basically one-shot snapshots. The thing I wanted was an interactive version of that, with the filters exposed, and with everything else I cared about layered on top.
The original-original idea was a small network of MEMS-microphone sensors on light poles streaming live ambient noise into the map. That’s still the long-term version. 311 turned out to be a surprisingly decent proxy in the meantime, and Open-Meteo gave me the microclimate piece for free. Riley Walz’s bop spotter is the spiritual cousin on the audio side, and Mr. Chilly’s distributed weather stations are basically the same idea for fog and temperature. The vibe I want this project to have is somewhere in that lineage: a little bit prankster, a little bit hacker, taking public data more seriously than the people publishing it.