diff --git a/notebooks/nighttime-lights/analysis-2024/.Rhistory b/notebooks/nighttime-lights/analysis-2024/.Rhistory
new file mode 100644
index 0000000..3857e58
--- /dev/null
+++ b/notebooks/nighttime-lights/analysis-2024/.Rhistory
@@ -0,0 +1,507 @@
+# Myanmar: 2024 Update
+# Filepaths --------------------------------------------------------------------
+db_dir <- file.path("~", "Dropbox", "World Bank", "Side Work", "Myanmar Economic Monitor")
+data_dir      <- file.path(db_dir, "data")
+ntl_bm_dir    <- file.path(data_dir, "Nighttime Lights BlackMarble")
+gas_flare_dir <- file.path(data_dir, "Global Gas Flaring 2023")
+gadm_dir      <- file.path(data_dir, "GADM")
+gdp_dir       <- file.path(data_dir, "GDP")
+wdi_dir       <- file.path(data_dir, "WDI")
+# Packages ---------------------------------------------------------------------
+library(dplyr)
+library(readr)
+library(janitor)
+library(raster)
+library(sf)
+library(stringr)
+library(readxl)
+library(exactextractr)
+#library(blackmarbler)
+library(geodata)
+library(lubridate)
+library(rgeos)
+library(leaflet)
+library(leaflet.extras)
+library(ggplot2)
+library(ggpubr)
+library(tidyr)
+library(purrr)
+library(stargazer)
+library(furrr)
+library(stringr)
+library(rhdf5)
+library(raster)
+library(dplyr)
+library(sf)
+library(lubridate)
+library(WDI)
+library(broom)
+library(blackmarbler)
+# Aggregate Nighttime Lights
+# Do if make changes to how data processed and need to start from scratch
+DELETE_DIR <- F
+if(DELETE_DIR){
+unlink(file.path(ntl_bm_dir, "FinalData", "aggregated"), recursive = T)
+dir.create(file.path(ntl_bm_dir, "FinalData", "aggregated"))
+}
+# Prep SEZ ---------------------------------------------------------------------
+sez_sf <- read_sf(file.path(data_dir, "SEZ", "RawData", "industrial__special_economic_zones_sept2019.shp"))
+sez_sf <- sez_sf %>% st_buffer(dist = 2.5)
+# Loop through ROIs ------------------------------------------------------------
+for(adm_level in rev(c("bound1", "bound2", "sez", "0", "1", "2", "3"))){
+if(adm_level == "sez"){
+roi_sf <- sez_sf
+} else if(adm_level == "bound1"){
+roi_sf <- read_sf(file.path(data_dir, "Boundaries", "RawData", "mmr_polbnda_adm1_250k_mimu.shp"))
+} else if(adm_level == "bound2"){
+roi_sf <- read_sf(file.path(data_dir, "Boundaries", "RawData", "mmr_polbnda_adm2_250k_mimu_1.shp"))
+} else{
+roi_sf <- read_sf(file.path(gadm_dir, "rawdata", paste0("gadm41_MMR_",adm_level,".json")))
+}
+# Loop through product -------------------------------------------------------
+# VNP46A2 = daily
+# VNP46A3 = monthly
+# VNP46A4 = annually
+for(product in c("VNP46A4", "VNP46A3")){
+## Make directory to export files - organized by ROI and prduct
+OUT_DIR <- file.path(ntl_bm_dir, "FinalData", "aggregated", paste0("adm", adm_level, "_", product))
+dir.create(OUT_DIR)
+# Loop through rasters -----------------------------------------------------
+r_name_vec <- file.path(ntl_bm_dir, "FinalData", paste0(product, "_rasters")) %>% list.files()
+for(r_name_i in r_name_vec){
+OUT_FILE <- file.path(OUT_DIR, r_name_i %>% str_replace_all(".tif", ".Rds"))
+## Check if file exists
+if(!file.exists(OUT_FILE)){
+## Load raster and create rasters for just gas flaring and non gas flaring locations
+r <- raster(file.path(ntl_bm_dir, "FinalData", paste0(product, "_rasters"), r_name_i))
+## Extract data
+ntl_df <- exact_extract(r, roi_sf, fun = c("mean", "median", "sum"))
+roi_sf$ntl_bm_mean   <- ntl_df$mean
+roi_sf$ntl_bm_median <- ntl_df$median
+roi_sf$ntl_bm_sum    <- ntl_df$sum
+## Prep for export
+roi_df <- roi_sf %>%
+st_drop_geometry()
+## Add date
+if(product == "VNP46A2"){
+year <- r_name_i %>% substring(12,15) %>% as.numeric()
+day  <- r_name_i %>% substring(12,21)
+date_r <- r_name_i %>%
+str_replace_all("VNP46A2_t", "") %>%
+str_replace_all(".tif", "") %>%
+str_replace_all("_", "-") %>%
+paste0("-01") %>%
+ymd()
+}
+if(product == "VNP46A3"){
+date_r <- r_name_i %>%
+str_replace_all("VNP46A3_t", "") %>%
+str_replace_all(".tif", "") %>%
+str_replace_all("_", "-") %>%
+paste0("-01") %>%
+ymd()
+}
+if(product == "VNP46A4"){
+# Just grab year
+date_r <- r_name_i %>%
+str_replace_all("VNP46A4_t", "") %>%
+str_replace_all(".tif", "") %>%
+as.numeric()
+}
+roi_df$date <- date_r
+## Export
+saveRDS(roi_df, OUT_FILE)
+}
+}
+}
+}
+# PSV Rider Feedback
+# Settings ---------------------------------------------------------------------
+#### CODE TO RUN
+RUN_CODE <- F # If F, doesn't run any code, no matter the below settings
+PROCESS_SENSOR_DATA        <- F # Download and clean sensor data (takes a while)
+PROCESS_SENSOR_SURVEY_DATA <- F # Clean survey data. RUN STATA MASTER FIRST TO DOWNLOAD DATA.
+#### Other
+# Relevant for sensor data. If F, checks if data has already been downloaded. If
+# already downloaded, skips downloading it. Set to T if want to redownload data.
+OVERWRITE_EXTRACTED_DATA <- F
+# Authenticate Google Sheets ---------------------------------------------------
+if(F){
+# Code for processing sensor survey data uploads some data to google sheets
+# (so the sensor company can see issues with the survey that they can correct).
+# Can run these two lines to make sure google sheets in authenticated.
+# Relevant for: 03_check_all_surveys_entered_compare_wialon.R
+library(googlesheets4)
+gs4_auth()
+}
+# Filepaths --------------------------------------------------------------------
+#### Root
+# Rob (Personal Comptuer)
+if(Sys.info()[["user"]] == "robmarty"){
+dropbox_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback"
+#db_pii_dir  <- "/Volumes/PSV_Rider_Feedback_PII"
+db_pii_dir <- "~/Documents/World Bank/PSV Rider Feedback"
+#onedrive_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback"
+github_dir <- "~/Documents/Github/PSV-Rider-Feedback/"
+}
+# Rob (WB Computer)
+if(Sys.info()[["user"]] == "WB521633"){
+dropbox_dir <- "C:/Users/wb521633/Dropbox/World Bank/IEs/PSV Rider Feedback"
+github_dir <- "C:/Users/wb521633/Documents/Github/PSV-Rider-Feedback/"
+onedrive_dir <- "C:/Users/wb521633/OneDrive - WBG/PSV Rider Feedback"
+}
+# Tom (personal computer)
+if(Sys.info()[["user"]] == "Tom"){
+dropbox_dir <- "/Users/Tom/Dropbox/PSV Rider Feedback/"
+github_dir <- "~/Google Drive/My Drive/Google Drive (iamtomharris98@gmail.com)/DIME 2/PSV-Rider-Feedback/"
+}
+# Ruiwen (WB Computer)
+if(Sys.info()[["user"]] == "wb575963"){
+dropbox_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback"
+github_dir <- "C:/Users/wb575963/Github/PSV-Rider-Feedback/"
+onedrive_dir <- "C:/Users/wb575963/WBG/Robert Andrew Marty - PSV Rider Feedback"
+}
+#### From Dropbox
+data_dir                   <- file.path(dropbox_dir, "Data")
+rider_feedback_dir         <- file.path(data_dir, "Rider Feedback")
+sensors_dir                <- file.path(data_dir, "Sensor Data")
+sensor_install_survey_dir  <- file.path(data_dir, "Matatu Sensor Installation Survey")
+sticker_install_survey_dir <- file.path(data_dir, "Sticker Installation Survey")
+sacco_route_dir            <- file.path(data_dir, "Sacco Route Data")
+driver_message_dir         <- file.path(data_dir, "Send Messages to Drivers")
+matatu_data_dir            <- file.path(data_dir, "Matatu Data")
+ntsa_speed_dir             <- file.path(data_dir, "NTSA - Speed Data")
+ap_data_dir <- file.path(dropbox_dir, "Academic Paper/Data")
+brief_figures_dir <- file.path(dropbox_dir, "Policy Brief/figures")
+#### Encrypted Data
+data_pii_dir                   <- file.path(db_pii_dir, "Data")
+rider_feedback_pii_dir         <- file.path(data_pii_dir, "Rider Feedback")
+sticker_install_survey_pii_dir <- file.path(data_pii_dir, "Sticker Installation Survey")
+# API Keys ---------------------------------------------------------------------
+if(Sys.info()[["user"]] == "robmarty"){
+api_keys_df <- read.csv(file.path("~/Dropbox", "World Bank", "Webscraping", "Files for Server", "api_keys.csv"),
+stringsAsFactors = F)
+}
+# Packages ---------------------------------------------------------------------
+library(tidyverse)
+library(dplyr)
+library(lubridate)
+library(gridExtra)
+library(hrbrthemes)
+library(readxl)
+library(labelled)
+library(haven)
+library(stringr)
+library(rmarkdown)
+library(wesanderson)
+library(knitr)
+library(data.table)
+library(DT)
+library(syn)
+library(kableExtra)
+library(openssl)
+library(ggmap)
+library(lubridate)
+library(httr)
+library(jsonlite)
+library(arrow)
+library(furrr)
+library(geosphere)
+library(sf)
+library(sfheaders)
+library(googlesheets4)
+library(leaflet)
+library(progress)
+library(ggpubr)
+library(twilio)
+library(raster)
+library(magick)
+library(janitor)
+library(qrcode)
+library(scales)
+# Functions --------------------------------------------------------------------
+#source(file.path(github_dir, "DataWork", "Sensor Data", "01_clean_data", "_wialon_api_functions.R"))
+source(file.path(github_dir, "Functions", "load_sensortracing_raw.R"))
+source("https://raw.githubusercontent.com/ramarty/r_google_translate/main/r_google_translate.R")
+# Code -------------------------------------------------------------------------
+if(RUN_CODE){
+# Process Sticker Installation Survey ----------------------------------------
+sticker_install_survey_code_dir <- file.path(github_dir, "DataWork", "Sticker Installation Survey")
+source(file.path(sticker_install_survey_code_dir, "01_clean_data.R"))
+# Process Rider Feedback Data ------------------------------------------------
+rider_feedback_code_dir <- file.path(github_dir, "DataWork", "Rider Feedback")
+source(file.path(rider_feedback_code_dir, "01_clean_data.R"))
+source(file.path(rider_feedback_code_dir, "02a_translate_comments.R"))
+source(file.path(rider_feedback_code_dir, "02b_append_translated_comments.R"))
+source(file.path(rider_feedback_code_dir, "03_merge_data.R"))
+source(file.path(rider_feedback_code_dir, "04_create_comments_to_code.R"))
+# Process Sensor Data --------------------------------------------------------
+if(PROCESS_SENSOR_DATA){
+sensor_code_dir <- file.path(github_dir, "DataWork", "Sensor Data", "01_clean_data")
+# Check Wailon Token Still Valid
+source(file.path(sensor_code_dir, "00_check_wialon_token.R"))
+# Download Sensor Data
+source(file.path(sensor_code_dir, "02_download_data.R"))
+# Process Sensor Tracing Data
+# A. Sensor data downloaded as json. Convert json to dataframe/export as
+#    parquet file.
+# B. Aggregate sensor data to hourly level and save file for each vehicle
+#    and day (so file at hourly level for each vehilce and day). Saves file
+#    with data including polyline of route and a separate smaller file
+#    that just includes the data (with no polyline)
+# C. Append sensor hourly data, creating file with all data appended.
+source(file.path(sensor_code_dir, "03a_sensor_data_json_to_df.R"))
+source(file.path(sensor_code_dir, "03b_sensor_to_hourly_individual_files.R"))
+source(file.path(sensor_code_dir, "03c_append_sensor_hr_data.R"))
+# Process Echo driving Data
+source(file.path(sensor_code_dir, "04a_append_echodriving_raw.R"))
+source(file.path(sensor_code_dir, "04a_echodriving_to_hourly_individual_files.R"))
+source(file.path(sensor_code_dir, "04b_append_echodriving_hr_data.R"))
+# Process Speed Data
+source(file.path(sensor_code_dir, "05a_append_speedings_raw.R"))
+source(file.path(sensor_code_dir, "05b_speedings_to_hr.R"))
+# Create Day and Day/Hour data with sensor tracing + echo driving merged in
+source(file.path(sensor_code_dir, "06_merge_dayhr_data.R"))
+source(file.path(sensor_code_dir, "07_aggregate_to_daily.R"))
+}
+# Process Sensor Installation Survey -----------------------------------------
+if(PROCESS_SENSOR_SURVEY_DATA){
+sensor_survey_code_dir <- file.path(github_dir, "DataWork", "Matatu Sensor Installation Survey")
+source(file.path(sensor_survey_code_dir, "02_clean_data.R"))
+source(file.path(sensor_survey_code_dir, "03_check_all_surveys_entered_compare_wialon.R"))
+}
+}
+# Crowdsourcing Safety Paper
+RUN_CODE <- F
+# Filepaths --------------------------------------------------------------------
+# Rob (Personal Comptuer)
+if(Sys.info()[["user"]] == "robmarty"){
+db_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback"
+data_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback/Academic Paper/Data"
+git_dir  <- "~/Documents/Github/PSV-Rider-Feedback/Academic Paper"
+overleaf_dir <- "~/Dropbox/Apps/Overleaf/Crowdsourcing Safety Kenya Matatu Passenger Safety Project"
+}
+# Ruiwen (WB Computer)
+if(Sys.info()[["user"]] == "wb575963"){
+db_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback"
+data_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback/Academic Paper/Data"
+git_dir <- "C:/Users/wb575963/Github/PSV-Rider-Feedback/Academic Paper"
+}
+tables_dir  <- file.path(overleaf_dir, "tables")
+figures_dir <- file.path(overleaf_dir, "figures")
+# Packages ---------------------------------------------------------------------
+library(magrittr)
+library(dplyr)
+library(readr)
+library(tidyr)
+library(stringr)
+library(lubridate)
+library(ggplot2)
+library(sf)
+library(readxl)
+library(janitor)
+library(arrow)
+library(quanteda)
+library(wordcloud2)
+library(tidytext)
+library(sentimentr)
+library(ggpubr)
+library(DescTools)
+library(forcats)
+library(tm)
+library(quanteda)
+library(scales)
+library(haven)
+library(fixest)
+library(purrr)
+library(tibble)
+library(did)
+#library(cleanNLP)
+#library(coreNLP)
+# Parameters -------------------------------------------------------------------
+DRIVING_WORDS <- "safe|drunk|accident|careless"
+COVID_WORDS <- "covid|pandemic|mask|social distance|sanitizer|sanitiser"
+# Run Code =====================================================================
+if(RUN_CODE){
+# Clean data -----------------------------------------------------------------
+git_clean_data_dir <- file.path(git_dir, "01_clean_data")
+source(file.path(git_clean_data_dir, "01_feedback_outliers.R"))
+source(file.path(git_clean_data_dir, "02_classify_feedback.R"))
+source(file.path(git_clean_data_dir, "03_make_vehicle_level_data.R"))
+# Passenger feedback analysis ------------------------------------------------
+git_feedback_dir <- file.path(git_dir, "02_analysis", "passenger_feedback_analysis")
+source(file.path(git_feedback_dir, "ratings_sum_stats_table.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_indiv_boxplot.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_vehicle_scatter.R"))
+source(file.path(git_feedback_dir, "safety_speed_v1_crosstab.R"))
+source(file.path(git_feedback_dir, "safety_speed_v2_crosstab.R"))
+source(file.path(git_feedback_dir, "sentiment_distribution.R"))
+source(file.path(git_feedback_dir, "sentiment_vs_class.R"))
+source(file.path(git_feedback_dir, "top_pos_neg_words.R"))
+source(file.path(git_feedback_dir, "vehicle_indicators_distribution.R"))
+# Telematics data analysis ---------------------------------------------------
+git_telematics_dir <- file.path(git_dir, "02_analysis", "telematics_analysis")
+source(file.path(git_telematics_dir, "compare_variables.R"))
+source(file.path(git_telematics_dir, "indicator_distribution.R"))
+source(file.path(git_telematics_dir, "indicator_sum_stat_table.R"))
+# Telematics vs feedback analysis --------------------------------------------
+git_tele_vs_feed_dir <- file.path(git_dir, "02_analysis", "telematics_vs_feedback")
+source(file.path(git_tele_vs_feed_dir, "correlation_plot.R"))
+source(file.path(git_tele_vs_feed_dir, "scatter_plots.R"))
+}
+# Crowdsourcing Safety Paper
+RUN_CODE <- F
+# Filepaths --------------------------------------------------------------------
+# Rob (Personal Comptuer)
+if(Sys.info()[["user"]] == "robmarty"){
+db_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback"
+data_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback/Academic Paper/Data"
+git_dir  <- "~/Documents/Github/PSV-Rider-Feedback/Academic Paper"
+overleaf_dir <- "~/Dropbox/Apps/Overleaf/Crowdsourcing Safety Kenya Matatu Passenger Safety Project"
+}
+# Ruiwen (WB Computer)
+if(Sys.info()[["user"]] == "wb575963"){
+db_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback"
+data_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback/Academic Paper/Data"
+git_dir <- "C:/Users/wb575963/Github/PSV-Rider-Feedback/Academic Paper"
+}
+tables_dir  <- file.path(overleaf_dir, "tables")
+figures_dir <- file.path(overleaf_dir, "figures")
+# Packages ---------------------------------------------------------------------
+library(magrittr)
+library(dplyr)
+library(readr)
+library(tidyr)
+library(stringr)
+library(lubridate)
+library(ggplot2)
+library(sf)
+library(readxl)
+library(janitor)
+library(arrow)
+library(quanteda)
+library(wordcloud2)
+library(tidytext)
+library(sentimentr)
+library(ggpubr)
+library(DescTools)
+library(forcats)
+library(tm)
+library(quanteda)
+library(scales)
+library(haven)
+library(fixest)
+library(purrr)
+library(tibble)
+library(did)
+#library(cleanNLP)
+#library(coreNLP)
+# Parameters -------------------------------------------------------------------
+DRIVING_WORDS <- "safe|drunk|accident|careless"
+COVID_WORDS <- "covid|pandemic|mask|social distance|sanitizer|sanitiser"
+# Run Code =====================================================================
+if(RUN_CODE){
+# Clean data -----------------------------------------------------------------
+git_clean_data_dir <- file.path(git_dir, "01_clean_data")
+source(file.path(git_clean_data_dir, "01_feedback_outliers.R"))
+source(file.path(git_clean_data_dir, "02_classify_feedback.R"))
+source(file.path(git_clean_data_dir, "03_make_vehicle_level_data.R"))
+# Passenger feedback analysis ------------------------------------------------
+git_feedback_dir <- file.path(git_dir, "02_analysis", "passenger_feedback_analysis")
+source(file.path(git_feedback_dir, "ratings_sum_stats_table.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_indiv_boxplot.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_vehicle_scatter.R"))
+source(file.path(git_feedback_dir, "safety_speed_v1_crosstab.R"))
+source(file.path(git_feedback_dir, "safety_speed_v2_crosstab.R"))
+source(file.path(git_feedback_dir, "sentiment_distribution.R"))
+source(file.path(git_feedback_dir, "sentiment_vs_class.R"))
+source(file.path(git_feedback_dir, "top_pos_neg_words.R"))
+source(file.path(git_feedback_dir, "vehicle_indicators_distribution.R"))
+# Telematics data analysis ---------------------------------------------------
+git_telematics_dir <- file.path(git_dir, "02_analysis", "telematics_analysis")
+source(file.path(git_telematics_dir, "compare_variables.R"))
+source(file.path(git_telematics_dir, "indicator_distribution.R"))
+source(file.path(git_telematics_dir, "indicator_sum_stat_table.R"))
+# Telematics vs feedback analysis --------------------------------------------
+git_tele_vs_feed_dir <- file.path(git_dir, "02_analysis", "telematics_vs_feedback")
+source(file.path(git_tele_vs_feed_dir, "correlation_plot.R"))
+source(file.path(git_tele_vs_feed_dir, "scatter_plots.R"))
+}
+# Crowdsourcing Safety Paper
+RUN_CODE <- F
+# Filepaths --------------------------------------------------------------------
+# Rob (Personal Comptuer)
+if(Sys.info()[["user"]] == "robmarty"){
+db_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback"
+data_dir <- "~/Dropbox/World Bank/IEs/PSV Rider Feedback/Academic Paper/Data"
+git_dir  <- "~/Documents/Github/PSV-Rider-Feedback/Academic Paper"
+overleaf_dir <- "~/Dropbox/Apps/Overleaf/Crowdsourcing Safety Kenya Matatu Passenger Safety Project"
+}
+# Ruiwen (WB Computer)
+if(Sys.info()[["user"]] == "wb575963"){
+db_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback"
+data_dir <- "C:/Users/wb575963/Dropbox/PSV Rider Feedback/Academic Paper/Data"
+git_dir <- "C:/Users/wb575963/Github/PSV-Rider-Feedback/Academic Paper"
+}
+tables_dir  <- file.path(overleaf_dir, "tables")
+figures_dir <- file.path(overleaf_dir, "figures")
+# Packages ---------------------------------------------------------------------
+library(magrittr)
+library(dplyr)
+library(readr)
+library(tidyr)
+library(stringr)
+library(lubridate)
+library(ggplot2)
+library(sf)
+library(readxl)
+library(janitor)
+library(arrow)
+library(quanteda)
+library(wordcloud2)
+library(tidytext)
+library(sentimentr)
+library(ggpubr)
+library(DescTools)
+library(forcats)
+library(tm)
+library(quanteda)
+library(scales)
+library(haven)
+library(fixest)
+library(purrr)
+library(tibble)
+library(did)
+#library(cleanNLP)
+#library(coreNLP)
+# Parameters -------------------------------------------------------------------
+DRIVING_WORDS <- "safe|drunk|accident|careless"
+COVID_WORDS <- "covid|pandemic|mask|social distance|sanitizer|sanitiser"
+# Run Code =====================================================================
+if(RUN_CODE){
+# Clean data -----------------------------------------------------------------
+git_clean_data_dir <- file.path(git_dir, "01_clean_data")
+source(file.path(git_clean_data_dir, "01_feedback_outliers.R"))
+source(file.path(git_clean_data_dir, "02_classify_feedback.R"))
+source(file.path(git_clean_data_dir, "03_make_vehicle_level_data.R"))
+# Passenger feedback analysis ------------------------------------------------
+git_feedback_dir <- file.path(git_dir, "02_analysis", "passenger_feedback_analysis")
+source(file.path(git_feedback_dir, "ratings_sum_stats_table.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_indiv_boxplot.R"))
+source(file.path(git_feedback_dir, "ratings_vs_comments_vehicle_scatter.R"))
+source(file.path(git_feedback_dir, "safety_speed_v1_crosstab.R"))
+source(file.path(git_feedback_dir, "safety_speed_v2_crosstab.R"))
+source(file.path(git_feedback_dir, "sentiment_distribution.R"))
+source(file.path(git_feedback_dir, "sentiment_vs_class.R"))
+source(file.path(git_feedback_dir, "top_pos_neg_words.R"))
+source(file.path(git_feedback_dir, "vehicle_indicators_distribution.R"))
+# Telematics data analysis ---------------------------------------------------
+git_telematics_dir <- file.path(git_dir, "02_analysis", "telematics_analysis")
+source(file.path(git_telematics_dir, "compare_variables.R"))
+source(file.path(git_telematics_dir, "indicator_distribution.R"))
+source(file.path(git_telematics_dir, "indicator_sum_stat_table.R"))
+# Telematics vs feedback analysis --------------------------------------------
+git_tele_vs_feed_dir <- file.path(git_dir, "02_analysis", "telematics_vs_feedback")
+source(file.path(git_tele_vs_feed_dir, "correlation_plot.R"))
+source(file.path(git_tele_vs_feed_dir, "scatter_plots.R"))
+}
diff --git a/notebooks/nighttime-lights/analysis-2024/01_process_data/02_download_ntl.R b/notebooks/nighttime-lights/analysis-2024/01_process_data/02_download_ntl.R
index 4abb77c..ce343a6 100644
--- a/notebooks/nighttime-lights/analysis-2024/01_process_data/02_download_ntl.R
+++ b/notebooks/nighttime-lights/analysis-2024/01_process_data/02_download_ntl.R
@@ -11,7 +11,7 @@ roi_sf <- read_sf(file.path(gadm_dir, "rawdata", paste0("gadm41_MMR_",0,".json")
 # Download ---------------------------------------------------------------------
 bm_raster(roi_sf = roi_sf,
           product_id = "VNP46A4",
-          date = 2012:2022,
+          date = 2012:2023,
           bearer = bearer,
           output_location_type = "file",
           file_dir = file.path(ntl_bm_dir, "FinalData", "VNP46A4_rasters"))
diff --git a/notebooks/nighttime-lights/analysis-2024/01_process_data/03_aggregate.R b/notebooks/nighttime-lights/analysis-2024/01_process_data/03_aggregate.R
index 2d4c9be..b32ef98 100644
--- a/notebooks/nighttime-lights/analysis-2024/01_process_data/03_aggregate.R
+++ b/notebooks/nighttime-lights/analysis-2024/01_process_data/03_aggregate.R
@@ -10,10 +10,17 @@ if(DELETE_DIR){
 
 # Prep SEZ ---------------------------------------------------------------------
 sez_sf <- read_sf(file.path(data_dir, "SEZ", "RawData", "industrial__special_economic_zones_sept2019.shp"))
-sez_sf <- sez_sf %>% st_buffer(dist = 2.5)
+sez_sf <- sez_sf %>% st_buffer(dist = 2500)
+
+border_df <- read_xlsx(file.path(data_dir, "Border Towns", "RawData", "Myanmar's Border Town Coordinates.xlsx"))
+border_sf <- border_df %>%
+  st_as_sf(coords = c("longitude", "latitude"), crs = 4326)
+
+border_1km_sf <- border_sf %>% st_buffer(dist = 1000)
+border_2_5km_sf <- border_sf %>% st_buffer(dist = 2500)
 
 # Loop through ROIs ------------------------------------------------------------
-for(adm_level in c("bound1", "bound2", "sez", "0", "1", "2", "3")){
+for(adm_level in c("bound1", "bound2", "sez", "0", "1", "2", "3", "border_1km", "border_2_5km")){
 
   if(adm_level == "sez"){
     roi_sf <- sez_sf
@@ -21,6 +28,10 @@ for(adm_level in c("bound1", "bound2", "sez", "0", "1", "2", "3")){
     roi_sf <- read_sf(file.path(data_dir, "Boundaries", "RawData", "mmr_polbnda_adm1_250k_mimu.shp"))
   } else if(adm_level == "bound2"){
     roi_sf <- read_sf(file.path(data_dir, "Boundaries", "RawData", "mmr_polbnda_adm2_250k_mimu_1.shp"))
+  } else if(adm_level == "border_1km"){
+    roi_sf <- border_1km_sf
+  } else if(adm_level == "border_2_5km"){
+    roi_sf <- border_2_5km_sf
   } else{
     roi_sf <- read_sf(file.path(gadm_dir, "rawdata", paste0("gadm41_MMR_",adm_level,".json")))
   }
diff --git a/notebooks/nighttime-lights/analysis-2024/01_process_data/04_append.R b/notebooks/nighttime-lights/analysis-2024/01_process_data/04_append.R
index 2d61b49..b459597 100644
--- a/notebooks/nighttime-lights/analysis-2024/01_process_data/04_append.R
+++ b/notebooks/nighttime-lights/analysis-2024/01_process_data/04_append.R
@@ -3,7 +3,8 @@
 roi = "adm2"
 product <- "VNP46A3"
 
-for(roi in c("adm0", "adm1", "adm2", "adm3", "admsez", "admbound2", "admbound1")){ 
+for(roi in c("adm0", "adm1", "adm2", "adm3", "admsez", "admbound2", "admbound1", 
+             "admborder_1km", "admborder_2_5km")){ 
   for(product in c("VNP46A3", "VNP46A4")){
     
     df <- file.path(ntl_bm_dir, "FinalData", "aggregated", paste0(roi, "_", product)) %>%
diff --git a/notebooks/nighttime-lights/analysis-2024/02_analysis/annual_trends.R b/notebooks/nighttime-lights/analysis-2024/02_analysis/annual_trends.R
index 2be4cb0..571a888 100644
--- a/notebooks/nighttime-lights/analysis-2024/02_analysis/annual_trends.R
+++ b/notebooks/nighttime-lights/analysis-2024/02_analysis/annual_trends.R
@@ -2,71 +2,85 @@
 
 # Load data --------------------------------------------------------------------
 adm1_df <- readRDS(file.path(ntl_bm_dir, "FinalData", "aggregated",
-                      paste0("adm1", "_", "VNP46A3", ".Rds")))
+                             paste0("adm1", "_", "VNP46A4", ".Rds")))
 
 adm2_df <- readRDS(file.path(ntl_bm_dir, "FinalData", "aggregated",
-                             paste0("adm2", "_", "VNP46A3", ".Rds")))
+                             paste0("adm2", "_", "VNP46A4", ".Rds")))
 
 adm3_df <- readRDS(file.path(ntl_bm_dir, "FinalData", "aggregated",
-                             paste0("adm3", "_", "VNP46A3", ".Rds")))
+                             paste0("adm3", "_", "VNP46A4", ".Rds")))
 
 sez_df <- readRDS(file.path(ntl_bm_dir, "FinalData", "aggregated",
-                             paste0("admsez", "_", "VNP46A3", ".Rds")))
+                            paste0("admsez", "_", "VNP46A4", ".Rds")))
 
-# Filter -----------------------------------------------------------------------
-clean_data <- function(df){
-  df %>%
-    mutate(year = date %>% year(),
-           month = date %>% month()) %>%
-    filter(month != 12,
-           year >= 2019) 
-}
-
-adm1_df <- adm1_df %>% 
-  clean_data() %>%
-  group_by(year, NAME_1) %>%
-  dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
-  ungroup()
-
-adm2_df <- adm2_df %>% 
-  clean_data() %>%
-  group_by(year, NAME_1, NAME_2) %>%
-  dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
-  ungroup()
-
-adm3_df <- adm3_df %>% 
-  clean_data() %>%
-  group_by(year, NAME_1, NAME_2, NAME_3) %>%
-  dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
-  ungroup()
+border_df <- readRDS(file.path(ntl_bm_dir, "FinalData", "aggregated",
+                               paste0("admborder_1km", "_", "VNP46A4", ".Rds")))
 
-sez_df <- sez_df %>% 
-  clean_data() %>%
-  group_by(year, Name) %>%
-  dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
-  ungroup()
+# Filter -----------------------------------------------------------------------
+# clean_data <- function(df){
+#   df %>%
+#     mutate(year = date %>% year(),
+#            month = date %>% month()) %>%
+#     filter(month != 12,
+#            year >= 2019) 
+# }
+# 
+# adm1_df <- adm1_df %>% 
+#   clean_data() %>%
+#   group_by(year, NAME_1) %>%
+#   dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
+#   ungroup()
+# 
+# adm2_df <- adm2_df %>% 
+#   clean_data() %>%
+#   group_by(year, NAME_1, NAME_2) %>%
+#   dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
+#   ungroup()
+# 
+# adm3_df <- adm3_df %>% 
+#   clean_data() %>%
+#   group_by(year, NAME_1, NAME_2, NAME_3) %>%
+#   dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
+#   ungroup()
+# 
+# sez_df <- sez_df %>% 
+#   clean_data() %>%
+#   group_by(year, Name) %>%
+#   dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
+#   ungroup()
+# 
+# border_df <- border_df %>% 
+#   clean_data() %>%
+#   group_by(year, Name) %>%
+#   dplyr::summarise(across(c(ntl_bm_sum, ntl_bm_mean, ntl_bm_median), mean)) %>%
+#   ungroup()
 
 # Trends -----------------------------------------------------------------------
 sez_df %>%
-  ggplot(aes(x = year, y = ntl_bm_sum)) +
-  geom_line() +
-  facet_wrap(~Name)
+  ggplot(aes(x = date, y = ntl_bm_sum)) +
+  geom_col() +
+  facet_wrap(~Name, scales = "free_y")
+
+border_df %>%
+  ggplot(aes(x = date, y = ntl_bm_sum)) +
+  geom_col() +
+  facet_wrap(~border_town, scales = "free_y")
 
 adm1_df %>%
   filter(NAME_1 == "Yangon") %>%
-  ggplot(aes(x = year, y = ntl_bm_sum)) +
+  ggplot(aes(x = date, y = ntl_bm_sum)) +
   geom_col() 
 
 adm2_df %>%
   filter(NAME_1 == "Yangon") %>%
-  ggplot(aes(x = year, y = ntl_bm_sum)) +
+  ggplot(aes(x = date, y = ntl_bm_sum)) +
   geom_col() +
   facet_wrap(~NAME_2)
 
 adm3_df %>%
   mutate(NAME_3_clean = paste(NAME_2, NAME_3)) %>%
   filter(NAME_1 == "Yangon") %>%
-  ggplot(aes(x = year, y = ntl_bm_sum)) +
+  ggplot(aes(x = date, y = ntl_bm_sum)) +
   geom_col() +
   facet_wrap(~NAME_3_clean,
              scales = "free_y")
diff --git a/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_0_sp.rds b/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_0_sp.rds
new file mode 100644
index 0000000..73cbd10
Binary files /dev/null and b/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_0_sp.rds differ
diff --git a/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_1_sp.rds b/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_1_sp.rds
new file mode 100644
index 0000000..2b86f5d
Binary files /dev/null and b/notebooks/nighttime-lights/analysis-2024/gadm36_SYR_1_sp.rds differ
diff --git a/notebooks/nighttime-lights/analysis-2024/gadm36_TUR_0_sp.rds b/notebooks/nighttime-lights/analysis-2024/gadm36_TUR_0_sp.rds
new file mode 100644
index 0000000..21fc9c5
Binary files /dev/null and b/notebooks/nighttime-lights/analysis-2024/gadm36_TUR_0_sp.rds differ