arthurgailes

# Load necessary library
pacman::p_load(
  stringi, data.table, duckplyr, readr, dplyr, collapse, duckdb, dbplyr, bench,
  ggplot2
)

# Function to generate a dataframe chunk
generate_data_chunk <- function(num_rows = 1000, num_cols = 100) {
  # Generate numeric columns
  numeric_cols <- replicate(n = num_cols/2, expr = runif(num_rows, 1, 10000), simplify = FALSE)

arthurgailes

# Load necessary library
pacman::p_load(
  stringi, data.table, duckplyr, readr, dplyr, collapse, duckdb, dbplyr, bench,
  ggplot2
)

# Function to generate a dataframe chunk
generate_data_chunk <- function(num_rows = 1000, num_cols = 100) {
  # Generate numeric columns
  numeric_cols <- replicate(n = num_cols/2, expr = runif(num_rows, 1, 10000), simplify = FALSE)

arthurgailes

#' Fix broken Census Connecticut blocks post-2022
#' See: https://github.com/walkerke/tigris/issues/178
library(tigris)
library(sf)
library(dplyr)
library(readr)
library(testthat)

# load official CT crosswalk
# missing a few block ids, but tracts are a consistent 1:1 merge so use those

arthurgailes

# quick R script for seeing the trends in downloads for a package
library(cranlogs)
library(ggplot2)
library(collapse)
library(lubridate)

package <- "shiny"
# get downloads for a specific date
x <- cran_downloads(packages=shiny, from="2015-06-01", to="2023-08-14")
head(x)

arthurgailes

# list and unlist an sf object
library(collapse)
library(sf)
library(testthat)

nc <- st_read(system.file("shape/nc.shp", package="sf")) 

nc_list <- rsplit(nc, by = seq_len(nrow(nc)))

nc_collapse <- unlist2d(nc_list, idcols = FALSE, recursive = FALSE) |> 

arthurgailes

# Get all indexes, then drop and add before and after writing to table
# Speeds up writes
get_index_table <- function(con, tablename, schemaname) {
  index_df <- dbGetQuery(con, paste0(
    "SELECT indexname, indexdef FROM pg_indexes
    WHERE tablename = '", tablename, "' AND schemaname = '", schemaname, "'"))
  
  index_df <- subset(index_df, !grepl("[pf]key", indexname))
  
  index_df$indexname <- paste0(schemaname, ".", index_df$indexname)

arthurgailes

if(!require(pacman)) install.packages('pacman')
pacman::p_load(tigris, tidycensus, furrr, purrr, tictoc, ggplot2)

options(tigris_use_cache = FALSE) # make things equal between runs

state_names <- c(state.name, "District of Columbia")
names(state_names) <- state_names

# purrr: 
tictoc::tic()

arthurgailes

# Create census commute share in Portland
if(!require(pacman)) install.packages('pacman')
pacman::p_load(dplyr, tidycensus)

# https://data.census.gov/cedsci/table?q=means%20transportation&g=1600000US4159000&tid=ACSDT1Y2021.B08006
trans_table <- 'B08006' 

# note, no data for 2020, the 2005 data is corrupt :
#   the non-auto counts are off
years <- c(2006:2019, 2021)

arthurgailes

license: mit

arthurgailes

license: mit