From fc10f4cd8c06941c965f6509999c112f0fce87ee Mon Sep 17 00:00:00 2001
From: Jean Cochrane <jean.cochrane@cookcountyil.gov>
Date: Wed, 4 Dec 2024 23:21:47 +0000
Subject: [PATCH 1/2] Pull class codes from Athena instead of `ccao` package in
 land_nbhd_rate script

---
 .../ccao/ccao-land-land_nbhd_rate.R           | 293 +++++++++---------
 1 file changed, 149 insertions(+), 144 deletions(-)

diff --git a/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R b/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
index 22a88d5ab..34f1080ca 100644
--- a/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
+++ b/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
@@ -1,144 +1,149 @@
-library(arrow)
-library(aws.s3)
-library(dplyr)
-library(openxlsx)
-library(purrr)
-library(readr)
-library(snakecase)
-library(stringr)
-library(tidyr)
-source("utils.R")
-
-# This script retrieves and cleans land value spreadsheets provided by
-# the Valuations department and formats them for use in Athena
-AWS_S3_RAW_BUCKET <- Sys.getenv("AWS_S3_RAW_BUCKET")
-AWS_S3_WAREHOUSE_BUCKET <- Sys.getenv("AWS_S3_WAREHOUSE_BUCKET")
-input_bucket <- file.path(AWS_S3_RAW_BUCKET, "ccao", "land")
-output_bucket <- file.path(AWS_S3_WAREHOUSE_BUCKET, "ccao", "land")
-
-# Location of remote files
-remote_file_raw_nbhd_rate_2022 <- file.path(
-  input_bucket, "nbhd_rate", "2022.xlsx"
-)
-remote_file_raw_nbhd_rate_2023 <- file.path(
-  input_bucket, "nbhd_rate", "2023.xlsx"
-)
-remote_file_raw_nbhd_rate_2024 <- file.path(
-  input_bucket, "nbhd_rate", "2024.xlsx"
-)
-remote_file_warehouse_nbhd_rate <- file.path(
-  output_bucket, "land_nbhd_rate"
-)
-
-
-# Temp file to download workbook
-tmp_file_nbhd_rate_2022 <- tempfile(fileext = ".xlsx")
-tmp_file_nbhd_rate_2023 <- tempfile(fileext = ".xlsx")
-tmp_file_nbhd_rate_2024 <- tempfile(fileext = ".xlsx")
-
-# Grab the workbook from the raw S3 bucket
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2022,
-  file = tmp_file_nbhd_rate_2022
-)
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2023,
-  file = tmp_file_nbhd_rate_2023
-)
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2024,
-  file = tmp_file_nbhd_rate_2024
-)
-
-# List of regression classes
-class <- ccao::class_dict %>%
-  filter(regression_class) %>%
-  pull(class_code)
-
-# Load the raw workbooks, rename and clean up columns
-land_nbhd_rate_2022 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2022) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  select(
-    township_code = twp_number,
-    township_name = twp_name,
-    town_nbhd = twp_nbhd,
-    `2019` = `2019_rate`,
-    `2022` = `2022_rate`
-  ) %>%
-  pivot_longer(
-    c(`2019`, `2022`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(
-    across(c(township_code:town_nbhd, year), as.character),
-    town_nbhd = str_remove_all(town_nbhd, "-"),
-    land_rate_per_sqft = parse_number(land_rate_per_sqft)
-  ) %>%
-  expand_grid(class)
-
-land_nbhd_rate_2023 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2023) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  select(
-    town_nbhd = neighborhood_id,
-    `2020` = `2020_2_00_class_unit_price`,
-    `2023` = `2023_2_00_class_unit_price`
-  ) %>%
-  mutate(
-    town_nbhd = gsub("\\D", "", town_nbhd),
-    township_code = substr(town_nbhd, 1, 2),
-    township_name = ccao::town_convert(township_code)
-  ) %>%
-  relocate(c(township_code, township_name)) %>%
-  pivot_longer(
-    c(`2020`, `2023`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
-  expand_grid(class)
-
-land_nbhd_rate_2024 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2024) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  mutate(
-    town_nbhd = paste0(
-      township_code, str_pad(neighborhood, 3, side = "left", pad = "0")
-    )
-  ) %>%
-  select(
-    town_nbhd,
-    classes,
-    `2021` = `2021_unit_price`,
-    `2024` = `2024_unit_price`
-  ) %>%
-  mutate(
-    town_nbhd = gsub("\\D", "", town_nbhd),
-    township_code = substr(town_nbhd, 1, 2),
-    township_name = ccao::town_convert(township_code)
-  ) %>%
-  relocate(c(township_code, township_name)) %>%
-  pivot_longer(
-    c(`2021`, `2024`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
-  expand_grid(class) %>%
-  # 2024 contains bifurcated neighborhood land rates across class
-  filter(
-    !(classes == "all other regression classes" & class %in% c("210", "295")),
-    !(classes == "2-10s/2-95s" & !(class %in% c("210", "295")))
-  ) %>%
-  select(-classes)
-
-# Write the rates to S3, partitioned by year
-bind_rows(
-  land_nbhd_rate_2022,
-  land_nbhd_rate_2023,
-  land_nbhd_rate_2024
-) %>%
-  relocate(land_rate_per_sqft, .after = last_col()) %>%
-  group_by(year) %>%
-  arrow::write_dataset(
-    path = remote_file_warehouse_nbhd_rate,
-    format = "parquet",
-    hive_style = TRUE,
-    compression = "snappy"
-  )
+library(arrow)
+library(aws.s3)
+library(dplyr)
+library(noctua)
+library(openxlsx)
+library(purrr)
+library(readr)
+library(snakecase)
+library(stringr)
+library(tidyr)
+source("utils.R")
+
+# This script retrieves and cleans land value spreadsheets provided by
+# the Valuations department and formats them for use in Athena
+AWS_S3_RAW_BUCKET <- Sys.getenv("AWS_S3_RAW_BUCKET")
+AWS_S3_WAREHOUSE_BUCKET <- Sys.getenv("AWS_S3_WAREHOUSE_BUCKET")
+input_bucket <- file.path(AWS_S3_RAW_BUCKET, "ccao", "land")
+output_bucket <- file.path(AWS_S3_WAREHOUSE_BUCKET, "ccao", "land")
+
+AWS_ATHENA_CONN_NOCTUA <- dbConnect(noctua::athena(), rstudio_conn_tab = FALSE)
+
+# Location of remote files
+remote_file_raw_nbhd_rate_2022 <- file.path(
+  input_bucket, "nbhd_rate", "2022.xlsx"
+)
+remote_file_raw_nbhd_rate_2023 <- file.path(
+  input_bucket, "nbhd_rate", "2023.xlsx"
+)
+remote_file_raw_nbhd_rate_2024 <- file.path(
+  input_bucket, "nbhd_rate", "2024.xlsx"
+)
+remote_file_warehouse_nbhd_rate <- file.path(
+  output_bucket, "land_nbhd_rate"
+)
+
+
+# Temp file to download workbook
+tmp_file_nbhd_rate_2022 <- tempfile(fileext = ".xlsx")
+tmp_file_nbhd_rate_2023 <- tempfile(fileext = ".xlsx")
+tmp_file_nbhd_rate_2024 <- tempfile(fileext = ".xlsx")
+
+# Grab the workbook from the raw S3 bucket
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2022,
+  file = tmp_file_nbhd_rate_2022
+)
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2023,
+  file = tmp_file_nbhd_rate_2023
+)
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2024,
+  file = tmp_file_nbhd_rate_2024
+)
+
+# List of regression classes
+class <- dbGetQuery(
+  AWS_ATHENA_CONN_NOCTUA,
+  "SELECT class_code FROM ccao.class_dict WHERE regression_class"
+) %>%
+  pull(class_code)
+
+# Load the raw workbooks, rename and clean up columns
+land_nbhd_rate_2022 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2022) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  select(
+    township_code = twp_number,
+    township_name = twp_name,
+    town_nbhd = twp_nbhd,
+    `2019` = `2019_rate`,
+    `2022` = `2022_rate`
+  ) %>%
+  pivot_longer(
+    c(`2019`, `2022`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(
+    across(c(township_code:town_nbhd, year), as.character),
+    town_nbhd = str_remove_all(town_nbhd, "-"),
+    land_rate_per_sqft = parse_number(land_rate_per_sqft)
+  ) %>%
+  expand_grid(class)
+
+land_nbhd_rate_2023 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2023) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  select(
+    town_nbhd = neighborhood_id,
+    `2020` = `2020_2_00_class_unit_price`,
+    `2023` = `2023_2_00_class_unit_price`
+  ) %>%
+  mutate(
+    town_nbhd = gsub("\\D", "", town_nbhd),
+    township_code = substr(town_nbhd, 1, 2),
+    township_name = ccao::town_convert(township_code)
+  ) %>%
+  relocate(c(township_code, township_name)) %>%
+  pivot_longer(
+    c(`2020`, `2023`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
+  expand_grid(class)
+
+land_nbhd_rate_2024 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2024) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  mutate(
+    town_nbhd = paste0(
+      township_code, str_pad(neighborhood, 3, side = "left", pad = "0")
+    )
+  ) %>%
+  select(
+    town_nbhd,
+    classes,
+    `2021` = `2021_unit_price`,
+    `2024` = `2024_unit_price`
+  ) %>%
+  mutate(
+    town_nbhd = gsub("\\D", "", town_nbhd),
+    township_code = substr(town_nbhd, 1, 2),
+    township_name = ccao::town_convert(township_code)
+  ) %>%
+  relocate(c(township_code, township_name)) %>%
+  pivot_longer(
+    c(`2021`, `2024`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
+  expand_grid(class) %>%
+  # 2024 contains bifurcated neighborhood land rates across class
+  filter(
+    !(classes == "all other regression classes" & class %in% c("210", "295")),
+    !(classes == "2-10s/2-95s" & !(class %in% c("210", "295")))
+  ) %>%
+  select(-classes)
+
+# Write the rates to S3, partitioned by year
+bind_rows(
+  land_nbhd_rate_2022,
+  land_nbhd_rate_2023,
+  land_nbhd_rate_2024
+) %>%
+  relocate(land_rate_per_sqft, .after = last_col()) %>%
+  group_by(year) %>%
+  arrow::write_dataset(
+    path = remote_file_warehouse_nbhd_rate,
+    format = "parquet",
+    hive_style = TRUE,
+    compression = "snappy"
+  )

From 303be387d1d8359f5d21ed9360132069227f8943 Mon Sep 17 00:00:00 2001
From: Jean Cochrane <jean.cochrane@cookcountyil.gov>
Date: Thu, 5 Dec 2024 21:51:17 +0000
Subject: [PATCH 2/2] Convert land_nbhd_rate back to DOS file format

---
 .../ccao/ccao-land-land_nbhd_rate.R           | 298 +++++++++---------
 1 file changed, 149 insertions(+), 149 deletions(-)

diff --git a/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R b/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
index 34f1080ca..9aa97fa86 100644
--- a/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
+++ b/etl/scripts-ccao-data-warehouse-us-east-1/ccao/ccao-land-land_nbhd_rate.R
@@ -1,149 +1,149 @@
-library(arrow)
-library(aws.s3)
-library(dplyr)
-library(noctua)
-library(openxlsx)
-library(purrr)
-library(readr)
-library(snakecase)
-library(stringr)
-library(tidyr)
-source("utils.R")
-
-# This script retrieves and cleans land value spreadsheets provided by
-# the Valuations department and formats them for use in Athena
-AWS_S3_RAW_BUCKET <- Sys.getenv("AWS_S3_RAW_BUCKET")
-AWS_S3_WAREHOUSE_BUCKET <- Sys.getenv("AWS_S3_WAREHOUSE_BUCKET")
-input_bucket <- file.path(AWS_S3_RAW_BUCKET, "ccao", "land")
-output_bucket <- file.path(AWS_S3_WAREHOUSE_BUCKET, "ccao", "land")
-
-AWS_ATHENA_CONN_NOCTUA <- dbConnect(noctua::athena(), rstudio_conn_tab = FALSE)
-
-# Location of remote files
-remote_file_raw_nbhd_rate_2022 <- file.path(
-  input_bucket, "nbhd_rate", "2022.xlsx"
-)
-remote_file_raw_nbhd_rate_2023 <- file.path(
-  input_bucket, "nbhd_rate", "2023.xlsx"
-)
-remote_file_raw_nbhd_rate_2024 <- file.path(
-  input_bucket, "nbhd_rate", "2024.xlsx"
-)
-remote_file_warehouse_nbhd_rate <- file.path(
-  output_bucket, "land_nbhd_rate"
-)
-
-
-# Temp file to download workbook
-tmp_file_nbhd_rate_2022 <- tempfile(fileext = ".xlsx")
-tmp_file_nbhd_rate_2023 <- tempfile(fileext = ".xlsx")
-tmp_file_nbhd_rate_2024 <- tempfile(fileext = ".xlsx")
-
-# Grab the workbook from the raw S3 bucket
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2022,
-  file = tmp_file_nbhd_rate_2022
-)
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2023,
-  file = tmp_file_nbhd_rate_2023
-)
-aws.s3::save_object(
-  object = remote_file_raw_nbhd_rate_2024,
-  file = tmp_file_nbhd_rate_2024
-)
-
-# List of regression classes
-class <- dbGetQuery(
-  AWS_ATHENA_CONN_NOCTUA,
-  "SELECT class_code FROM ccao.class_dict WHERE regression_class"
-) %>%
-  pull(class_code)
-
-# Load the raw workbooks, rename and clean up columns
-land_nbhd_rate_2022 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2022) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  select(
-    township_code = twp_number,
-    township_name = twp_name,
-    town_nbhd = twp_nbhd,
-    `2019` = `2019_rate`,
-    `2022` = `2022_rate`
-  ) %>%
-  pivot_longer(
-    c(`2019`, `2022`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(
-    across(c(township_code:town_nbhd, year), as.character),
-    town_nbhd = str_remove_all(town_nbhd, "-"),
-    land_rate_per_sqft = parse_number(land_rate_per_sqft)
-  ) %>%
-  expand_grid(class)
-
-land_nbhd_rate_2023 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2023) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  select(
-    town_nbhd = neighborhood_id,
-    `2020` = `2020_2_00_class_unit_price`,
-    `2023` = `2023_2_00_class_unit_price`
-  ) %>%
-  mutate(
-    town_nbhd = gsub("\\D", "", town_nbhd),
-    township_code = substr(town_nbhd, 1, 2),
-    township_name = ccao::town_convert(township_code)
-  ) %>%
-  relocate(c(township_code, township_name)) %>%
-  pivot_longer(
-    c(`2020`, `2023`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
-  expand_grid(class)
-
-land_nbhd_rate_2024 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2024) %>%
-  set_names(snakecase::to_snake_case(names(.))) %>%
-  mutate(
-    town_nbhd = paste0(
-      township_code, str_pad(neighborhood, 3, side = "left", pad = "0")
-    )
-  ) %>%
-  select(
-    town_nbhd,
-    classes,
-    `2021` = `2021_unit_price`,
-    `2024` = `2024_unit_price`
-  ) %>%
-  mutate(
-    town_nbhd = gsub("\\D", "", town_nbhd),
-    township_code = substr(town_nbhd, 1, 2),
-    township_name = ccao::town_convert(township_code)
-  ) %>%
-  relocate(c(township_code, township_name)) %>%
-  pivot_longer(
-    c(`2021`, `2024`),
-    names_to = "year", values_to = "land_rate_per_sqft"
-  ) %>%
-  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
-  expand_grid(class) %>%
-  # 2024 contains bifurcated neighborhood land rates across class
-  filter(
-    !(classes == "all other regression classes" & class %in% c("210", "295")),
-    !(classes == "2-10s/2-95s" & !(class %in% c("210", "295")))
-  ) %>%
-  select(-classes)
-
-# Write the rates to S3, partitioned by year
-bind_rows(
-  land_nbhd_rate_2022,
-  land_nbhd_rate_2023,
-  land_nbhd_rate_2024
-) %>%
-  relocate(land_rate_per_sqft, .after = last_col()) %>%
-  group_by(year) %>%
-  arrow::write_dataset(
-    path = remote_file_warehouse_nbhd_rate,
-    format = "parquet",
-    hive_style = TRUE,
-    compression = "snappy"
-  )
+library(arrow)
+library(aws.s3)
+library(dplyr)
+library(noctua)
+library(openxlsx)
+library(purrr)
+library(readr)
+library(snakecase)
+library(stringr)
+library(tidyr)
+source("utils.R")
+
+# This script retrieves and cleans land value spreadsheets provided by
+# the Valuations department and formats them for use in Athena
+AWS_S3_RAW_BUCKET <- Sys.getenv("AWS_S3_RAW_BUCKET")
+AWS_S3_WAREHOUSE_BUCKET <- Sys.getenv("AWS_S3_WAREHOUSE_BUCKET")
+input_bucket <- file.path(AWS_S3_RAW_BUCKET, "ccao", "land")
+output_bucket <- file.path(AWS_S3_WAREHOUSE_BUCKET, "ccao", "land")
+
+AWS_ATHENA_CONN_NOCTUA <- dbConnect(noctua::athena(), rstudio_conn_tab = FALSE)
+
+# Location of remote files
+remote_file_raw_nbhd_rate_2022 <- file.path(
+  input_bucket, "nbhd_rate", "2022.xlsx"
+)
+remote_file_raw_nbhd_rate_2023 <- file.path(
+  input_bucket, "nbhd_rate", "2023.xlsx"
+)
+remote_file_raw_nbhd_rate_2024 <- file.path(
+  input_bucket, "nbhd_rate", "2024.xlsx"
+)
+remote_file_warehouse_nbhd_rate <- file.path(
+  output_bucket, "land_nbhd_rate"
+)
+
+
+# Temp file to download workbook
+tmp_file_nbhd_rate_2022 <- tempfile(fileext = ".xlsx")
+tmp_file_nbhd_rate_2023 <- tempfile(fileext = ".xlsx")
+tmp_file_nbhd_rate_2024 <- tempfile(fileext = ".xlsx")
+
+# Grab the workbook from the raw S3 bucket
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2022,
+  file = tmp_file_nbhd_rate_2022
+)
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2023,
+  file = tmp_file_nbhd_rate_2023
+)
+aws.s3::save_object(
+  object = remote_file_raw_nbhd_rate_2024,
+  file = tmp_file_nbhd_rate_2024
+)
+
+# List of regression classes
+class <- dbGetQuery(
+  AWS_ATHENA_CONN_NOCTUA,
+  "SELECT class_code FROM ccao.class_dict WHERE regression_class"
+) %>%
+  pull(class_code)
+
+# Load the raw workbooks, rename and clean up columns
+land_nbhd_rate_2022 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2022) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  select(
+    township_code = twp_number,
+    township_name = twp_name,
+    town_nbhd = twp_nbhd,
+    `2019` = `2019_rate`,
+    `2022` = `2022_rate`
+  ) %>%
+  pivot_longer(
+    c(`2019`, `2022`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(
+    across(c(township_code:town_nbhd, year), as.character),
+    town_nbhd = str_remove_all(town_nbhd, "-"),
+    land_rate_per_sqft = parse_number(land_rate_per_sqft)
+  ) %>%
+  expand_grid(class)
+
+land_nbhd_rate_2023 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2023) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  select(
+    town_nbhd = neighborhood_id,
+    `2020` = `2020_2_00_class_unit_price`,
+    `2023` = `2023_2_00_class_unit_price`
+  ) %>%
+  mutate(
+    town_nbhd = gsub("\\D", "", town_nbhd),
+    township_code = substr(town_nbhd, 1, 2),
+    township_name = ccao::town_convert(township_code)
+  ) %>%
+  relocate(c(township_code, township_name)) %>%
+  pivot_longer(
+    c(`2020`, `2023`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
+  expand_grid(class)
+
+land_nbhd_rate_2024 <- openxlsx::read.xlsx(tmp_file_nbhd_rate_2024) %>%
+  set_names(snakecase::to_snake_case(names(.))) %>%
+  mutate(
+    town_nbhd = paste0(
+      township_code, str_pad(neighborhood, 3, side = "left", pad = "0")
+    )
+  ) %>%
+  select(
+    town_nbhd,
+    classes,
+    `2021` = `2021_unit_price`,
+    `2024` = `2024_unit_price`
+  ) %>%
+  mutate(
+    town_nbhd = gsub("\\D", "", town_nbhd),
+    township_code = substr(town_nbhd, 1, 2),
+    township_name = ccao::town_convert(township_code)
+  ) %>%
+  relocate(c(township_code, township_name)) %>%
+  pivot_longer(
+    c(`2021`, `2024`),
+    names_to = "year", values_to = "land_rate_per_sqft"
+  ) %>%
+  mutate(across(c(township_code:town_nbhd, year), as.character)) %>%
+  expand_grid(class) %>%
+  # 2024 contains bifurcated neighborhood land rates across class
+  filter(
+    !(classes == "all other regression classes" & class %in% c("210", "295")),
+    !(classes == "2-10s/2-95s" & !(class %in% c("210", "295")))
+  ) %>%
+  select(-classes)
+
+# Write the rates to S3, partitioned by year
+bind_rows(
+  land_nbhd_rate_2022,
+  land_nbhd_rate_2023,
+  land_nbhd_rate_2024
+) %>%
+  relocate(land_rate_per_sqft, .after = last_col()) %>%
+  group_by(year) %>%
+  arrow::write_dataset(
+    path = remote_file_warehouse_nbhd_rate,
+    format = "parquet",
+    hive_style = TRUE,
+    compression = "snappy"
+  )