Author: David Faden
-
“Are Emergent Abilities of Large Language Models a Mirage?”
A couple thoughts looking at this fascinating paper from 2023: “Are Emergent Abilities of Large Language Models a Mirage?”:
-
Telling g++ about homebrew-installed lib
I’ve been having fun looking at some code created by a friend/ex-coworker, including https://github.com/hfgong/walker_vose. It makes use of the beloved to me GoogleTest C++ test framework. I guess probably the right thing to do is to set things up via one of the methods mentioned in the GoogleTest docs, Bazel or CMake. (I was very…
-
Undoing perspectival distortion
I have a lot of photos containing a sign or other rectangular object where I’d like to undo the distortion that comes from having the plane of the camera not exactly parallel to the plane of the rectangle. I’ve also been wanting to play with Tauri, a framework for creating apps, for a while. So,…
-
California cities from CDTFA
At least based on the dates listed on https://lab.data.ca.gov/dataset/california-city-boundaries-and-identifiers — still says “03/24/25” — I’m guessing the city boundaries haven’t been updated yet. For a current project, I want the best boundaries available so let’s work directly with the source pointed out by the CIO, GIS: https://gis.data.ca.gov/datasets/CDTFA::city-and-county-boundary-line-changes/explore?layer=0&location=33.924386%2C-118.008294%2C11.67. > library(sf)> d <- read_sf(“~/Downloads/City_and_County_Boundary_Line_Changes_217120214122415134.gpkg”) It turns out…
-
A hole in Los Altos (continued)
The city clerk got back to me this morning. (Thanks much for the prompt response!) The hole in the map is also a hole in the ground — a reservoir belonging to the City of Mountain View. We happened to be in the area after checking out Tom’s Depot Cafe yesterday so I snapped a…
-
A hole in Los Altos
Looking at the Census page for Los Altos, something funny caught my eye: there’s a hole in the map for Los Altos: https://data.census.gov/profile/Los_Altos_city,_California?g=160XX00US0643280 Returning to the data from https://lab.data.ca.gov/dataset/california-city-boundaries-and-identifiers, we see the same thing: > library(ggplot2)> library(sf)> cities <- read_sf(“~/Downloads/California_Cities_and_Identifiers_Blue_Version_view_-6943741225906831761.gpkg”)> losAltos <- subset(cities, CDTFA_CITY == “Los Altos”)> ggplot(losAltos) + geom_sf() What’s in that hole? >…
-
What Wittier Whittier
I was really impressed to get an almost immediate reply from the CIO, GIS to my note about Whittier appearing to some extra stuff in the geometries provided here: https://lab.data.ca.gov/dataset/california-city-boundaries-and-identifiers. They pointed to the California Department of Tax and Fee Administration as the original source of the geometries, which has been updated in the meantime:…
-
California cities (continued)
Let’s get back roughly to where we were with the R session from California cities, reading in a copy of the data pulled from Wikipedia. > library(dplyr)> library(fillpattern)> library(sf)> library(ggplot2)> cities <- read_sf(“~/Downloads/California_Cities_and_Identifiers_Blue_Version_view_-6943741225906831761.gpkg”)> stats <- read.csv(url(“https://raw.githubusercontent.com/fadend/ca_cities_data/refs/heads/main/ca_cities.csv”), stringsAsFactors=FALSE) Oops. There’s some disagreement on names: > setdiff(stats$city_name, cities$city_name)[1] “Angels Camp” “California City” “City of Industry”> setdiff(cities$city_name, stats$city_name)[1]…
-
Fun with Xorshift
I’ve been hacking on an old game I wrote for the school paper I founded: Hangman. (Code: https://github.com/fadend/hangman.) One feature I like: the ability to create your own games with a shareable link. Here’s one for you: https://revfad.com/hangman/?w=JVFUVAT+LBH+N+JBAQRESHY+QNL. Since I didn’t want the URL parameter to give away the answer, I have the game run…
-
California cities
Here’s something fun: boundaries for cities in California: https://lab.data.ca.gov/dataset/california-city-boundaries-and-identifiers. Working with this in R: > library(sf)> cities <- read_sf(“~/Downloads/California_Cities_and_Identifiers_Blue_Version_view_-6943741225906831761.gpkg”) There are a few columns with the city names; “CENSUS_PLACE_NAME” and “CDTFA_CITY” (CA Department of Tax and Fee Administration?) agree so let’s just use “CDTFA_CITY” > sum(cities$CENSUS_PLACE_NAME != cities$CDTFA_CITY)[1] 0 It’s nice to see that there…