Merge pull request #9 from sarahhurley/main

pzaldonis · web-flow · commit b975deffac7c · 2024-12-10T15:30:58.000-05:00
add neighboring towns script
diff --git a/december_demo/Find_neighboring_towns.R b/december_demo/Find_neighboring_towns.R
@@ -0,0 +1,93 @@
+# This script will create lists of neighboring towns in Connecticut.
+# Author: Sarah Hurley
+  
+# Packages
+library(tidyverse)
+library(sf)
+library(arcgis)
+
+# Install packages
+# install.packages("tidyverse")
+# install.packages("sf")
+# remotes::install_github("r-arcgis/arcgis", dependencies = TRUE)
+
+# ============================================================================
+
+# Read in CT Municipalities layer (ArcGIS REST URL)
+town_url <- "https://services1.arcgis.com/FCaUeJ5SOVtImake/arcgis/rest/services/CT_Municipalities/FeatureServer/0"
+towns <- arc_open(town_url) %>% arc_select()
+
+plot(towns["Municipality"])
+
+
+# Create a tiny buffer around each towns geometry so we can capture the neighbors
+towns2 <- st_buffer(towns, dist = 0.001)
+
+# Get a list of lists of the overlapping neighbors
+neighbors <- st_intersects(towns2,
+                           towns2)
+
+
+# A function returning data frame of neighbors of a given town
+nbr_pairs <- function(idx)  {
+  data.frame(Town = rep(towns$Municipality[idx], length(neighbors[[idx]])), 
+             Neighbor_town = towns$Municipality[neighbors[[idx]]])
+}
+
+
+# Apply the function to list of indices (the neighbors list)
+pairs_of_names <- purrr::map_dfr(seq_along(neighbors),
+                                 nbr_pairs)
+
+
+grouped_list <- pairs_of_names %>% 
+  mutate(same = if_else(Town == Neighbor_town, TRUE, FALSE)) %>% 
+  filter(same == FALSE) %>% 
+  select(-same) %>% 
+  group_by(Town) %>%
+  mutate(Neighbor_index = row_number()) %>%
+  ungroup() 
+
+
+
+## Create 3 new data sets from the grouped list
+
+# 1. Number of neighboring towns for each town
+summary_towns <- grouped_list %>% 
+  group_by(Town) %>% 
+  summarise(Count = n())
+
+# 2. Wide data set: columns are # of neighboring towns
+wide_list <- grouped_list %>%
+  pivot_wider(names_from = Neighbor_index, 
+              values_from = Neighbor_town, 
+              names_prefix = "Neighbor_") %>% 
+  left_join(summary_towns, by = "Town") %>%
+  select(Town, Count, everything())
+
+# 3. Long data set: two columns, one for town and another for a neighboring town
+long_list <- grouped_list %>% 
+  select(Town, Neighbor_town)
+
+
+
+
+# write_csv(wide_list, "CT_Adjacent_Towns_WideFormat.csv")
+# write_csv(long_list, "CT_Adjacent_Towns_LongList.csv")
+
+
+
+
+
+### Resources ###
+# 1. Connecticut Neighboring Towns data set
+## https://data.ct.gov/Local-Government/Connecticut-Neighboring-Towns/3u6x-c464/about_data
+
+# 2. Wide vs. Long data
+## "Getting wide and long with data" by Stephen Skalicky: https://tinyurl.com/musuvthk
+
+# 3. R-ArcGIS Packages
+## https://github.com/R-ArcGIS/arcgislayers
+
+
+
