#5

@velgaks ,  it was simple.  We should have used inner join

manipulation/1-ellis-old.R

@@ -0,0 +1,298 @@
+#' ---
+#' title: "Ellis Lane"
+#' author: 
+#'   - "Andriy Koval"
+#'   - "other author"
+#' date: "last Updated: `r Sys.Date()`"
+#' ---
+#+ echo=F
+# rmarkdown::render(input = "./manipulation/1-ellis.R") # run to knit, don't uncomment
+
+#+ echo=F ----------------------------------------------------------------------
+library(knitr)
+# align the root with the project working directory
+opts_knit$set(root.dir='../')  #Don't combine this call with any
+#+ echo=F ----------------------------------------------------------------------
+rm(list = ls(all.names = TRUE)) # Clear the memory of variables from previous run.
+#This is not called by knitr, because it's above the first chunk.
+#+ results="hide",echo=F -------------------------------------------------------
+cat("\014") # Clear the console
+#+ echo=FALSE, results="asis" --------------------------------------------------
+cat("Report's native working directory: `", getwd(),"`") # Must be set to Project Directory
+#+ echo=F, results="asis" ------------------------------------------------------
+cat("\n# 1.Environment")
+#+ set_options, echo=F ---------------------------------------------------------
+echo_chunks <- TRUE
+eval_chunks <- TRUE
+cache_chunks <- TRUE
+report_render_start_time <- Sys.time()
+options(width=100) # number of characters to display in the output (dflt = 80)
+Sys.setlocale("LC_CTYPE", "russian")
+#+ load-sources ------------------------------------------------------------
+base::source("./scripts/common-functions.R") # project-level
+#+ load-packages -----------------------------------------------------------
+# Prefer to be greedy: load only what's needed
+# Three ways, from least (1) to most(3) greedy:
+# -- 1.Attach these packages so their functions don't need to be qualified: http://r-pkgs.had.co.nz/namespace.html#search-path
+library(ggplot2)   # graphs
+library(tidyverse)
+# -- 2.Import only certain functions of a package into the search path.
+import::from("magrittr", "%>%")
+# -- 3. Verify these packages are available on the machine, but their functions need to be qualified: http://r-pkgs.had.co.nz/namespace.html#search-path
+library("dplyr"    )# Avoid attaching dplyr, b/c its function names conflict with a lot of packages (esp base, stats, and plyr).
+requireNamespace("readr"    )# data import/export
+requireNamespace("tidyr"    )# tidy data
+requireNamespace("janitor"  )# tidy data
+requireNamespace("forcats"  )# factors
+requireNamespace("stringr"  )# strings
+requireNamespace("lubridate")# dates
+
+#+ declare-globals -------------------------------------------------------------
+# Constant values that won't change throughout the report
+
+# printed figures will go here:
+prints_folder <- paste0("./manipulation/prints/")
+if (!fs::dir_exists(prints_folder)) { fs::dir_create(prints_folder) }
+
+
+names_rada <- c(
+  "oblast"
+  ,"rai_center"
+  ,"rada_name"
+  ,"rada_code"
+  ,"hromada_name"
+  ,"hromada_code"
+  ,"note"
+)
+
+names_hromada <- c(
+   "category"
+  ,"oblast"
+  ,"raion"
+  ,"hromada_name"
+  ,"hromada_code"
+  ,"main_rada_code"
+  ,"rada_codes_v1"
+  ,"decision_date_v1"
+  ,"rada_codes_v2"
+  ,"decision_date_v2"
+  ,"rada_codes_v3"
+  ,"decision_date_v3"
+  ,"rada_codes_v4"
+  ,"decision_date_v4"
+  ,"rada_codes_v5"
+  ,"decision_date_v5"
+  ,"rada_codes_final"
+  ,"voluntary_amalgamation"
+)
+
+names_event <- c(
+  "hromada_code"
+  ,"rada_codes_v1"   
+  ,"decision_date_v1"   
+  ,"rada_codes_v2"   
+  ,"decision_date_v2"   
+  ,"rada_codes_v3"   
+  ,"decision_date_v3"   
+  ,"rada_codes_v4"   
+  ,"decision_date_v4"   
+  ,"rada_codes_v5"   
+  ,"decision_date_v5"  
+)
+
+names_admin_codes <- 
+  c(
+    "level_1"
+    ,"level_2"
+    ,"level_3"
+    ,"level_4"
+    ,"level_additional"
+    ,"object_category"
+    ,"object_name"
+  )
+
+col_types_hromada_raw <- readr::cols_only(
+  `Hromada_category`          = readr::col_character(),
+  `Region`                    = readr::col_character(),
+  `Raion`                     = readr::col_character(),
+  `Hromada`                   = readr::col_character(),
+  `Hromada_code`              = readr::col_character(),
+  `Main_council_code`         = readr::col_character(),
+  `Council_codes_v1`          = readr::col_character(),
+  `Decision_date_v1`          = readr::col_character(),
+  `Council_codes_v2`          = readr::col_character(),
+  `Decision_date_v2`          = readr::col_character(),
+  `Council_codes_v3`          = readr::col_character(),
+  `Decision_date_v3`          = readr::col_character(),
+  `Council_codes_v4`          = readr::col_character(),
+  `Decision_date_v4`          = readr::col_character(),
+  `Council_codes_v5`          = readr::col_character(),
+  `Decision_date_v5`          = readr::col_character(),
+  `Council_codes_final`       = readr::col_character(),
+  `voluntary_amalgamation`    = readr::col_character()
+)
+
+path_rada <- "./data-private/raw/rada.csv"
+path_hromada <- "./data-private/raw/hromada.csv"
+path_admin <- "./data-private/raw/ua-admin-codes.csv"    
+#+ declare-functions -----------------------------------------------------------
+
+#+ results="asis", echo=F ------------------------------------------------------
+cat("\n# 2.Data ")
+#+ load-data, eval=eval_chunks -------------------------------------------------
+ds0_admin   <- readr::read_csv(path_admin, col_names = names_admin_codes,skip=1)
+
+ds0_rada    <- readr::read_csv(path_rada, col_names = names_rada, skip = 1)
+ds0_hromada <- readr::read_csv(
+  path_hromada
+  ,col_types = col_types_hromada_raw # because had some trouble with parsing issues
+  ,skip = 0
+)
+
+
+#+ inspect-data ----------------------------------------------------------------
+ds0_admin %>% glimpse()
+ds0_rada    %>% glimpse()
+ds0_hromada %>% glimpse()
+
+
+ds0_hromada %>% slice(1100) %>% t()
+
+
+#+ tweak-data, eval=eval_chunks ------------------------------------------------
+ds0_rada %>% glimpse()
+ds1_rada <- 
+  ds0_rada %>% 
+  select(
+    rada_code
+    ,rada_name
+    ,hromada_code
+    # ,rai_center # not sure we need this
+    # ,oblast
+    # ,hromada_name
+  ) %>% 
+  arrange(
+    # oblast
+    rada_code
+    # ,rai_center # not sure we need this
+  )
+ds1_rada %>% glimpse()
+
+ds0_hromada %>% glimpse()
+names(ds0_hromada) <- names_hromada
+
+ds1_hromada <-
+  ds0_hromada %>% 
+  select(!starts_with("rada_codes_v")) %>% 
+  select(!starts_with("decision_date")) %>% 
+  select(
+    hromada_code
+    ,hromada_name
+    ,main_rada_code
+    ,rada_codes_final
+  )
+ds1_hromada %>% glimpse()  
+
+
+ds1_event <- 
+  ds0_hromada %>% 
+  select(names_event)
+
+
+ds1_rada %>% glimpse()
+ds1_hromada %>% glimpse()
+ds1_event %>% glimpse()
+
+
+#+ make-tidy-data --------------------------------------------------------------
+#+ 
+#+ 
+selected_codes <- c("UA05020010000053508","UA05020030000031457" )
+
+ds2 <- 
+  ds1_event %>% 
+  # filter(hromada_code %in% selected_codes) %>% 
+  tidyr::pivot_longer(
+    cols = !starts_with("hromada_code")
+  ) %>% 
+  mutate(
+    wave  = str_extract(name, "\\d$" )
+    ,name = str_remove( name, "^rada_|decision_")
+    ,name = str_remove( name, "_v\\d{1}$")
+  ) %>% 
+  pivot_wider(
+    names_from = "name", values_from = "value"
+  ) %>% 
+  mutate(
+    rada_code = str_split(codes, ',')
+  ) %>% 
+  unnest(cols = c("rada_code")) %>% 
+  select(-codes, -wave) %>% 
+  filter(!is.na(date)) %>%
+  mutate(
+    date = as.Date(date)
+    # ,rada_code = as.integer(rada_code)
+  ) %>% 
+  filter(!is.na(rada_code)) %>% 
+  arrange(hromada_code, date, rada_code)
+
+
+
+#+ table-1 ---------------------------------------------------------------------
+ds2 %>% glimpse()
+d <-
+  ds2  %>% 
+  group_by(hromada_code, date) %>% 
+  summarize(
+    rada_count = n_distinct(rada_code, na.rm = T)
+    ,.groups = "drop"
+  ) %>% 
+  left_join(
+    ds_map
+    ,by = "hromada_code"
+  )
+d
+
+g <-
+  d %>% 
+  filter(!is.na(hromada_code)) %>% 
+  ggplot(aes(
+    x=date
+    , y = rada_count
+    , group = hromada_code
+    # ,color = oblast_name
+    )
+  )+
+  geom_point(shape=21, alpha = .4, size = 1)+
+  facet_wrap(facets = c("oblast_name"))+
+  geom_line(alpha = .3)+
+  theme(
+   legend.position = "none"
+  )+
+  labs(
+    title = "Динаміка складу територіальних громад"
+    ,subtitle = "Як змінювалась кількість місцевих рад у складі громад?"
+    ,y = "Кількість місцевих рад у громаді"
+    ,x = "Дата зміни складу громад"
+  )
+
+g %>% 
+  quick_save("hromada-growth-n-rada",w=12, h=8)
+
+#+ graph-1 ---------------------------------------------------------------------
+#+ graph-2 ---------------------------------------------------------------------
+#+ save-to-disk, eval=eval_chunks-----------------------------------------------
+
+#+ results="asis", echo=F ------------------------------------------------------
+cat("\n# A. Session Information{#session-info}")
+#+ results="show", echo=F ------------------------------------------------------
+#' For the sake of documentation and reproducibility, the current report was rendered in the following environment.
+if( requireNamespace("devtools", quietly = TRUE) ) {
+  devtools::session_info()
+} else {
+  sessionInfo()
+}
+report_render_duration_in_seconds <- scales::comma(as.numeric(difftime(Sys.time(), report_render_start_time, units="secs")),accuracy=1)
+report_render_duration_in_minutes <- scales::comma(as.numeric(difftime(Sys.time(), report_render_start_time, units="mins")),accuracy=1)
+#' Report rendered by `r Sys.info()["user"]` at `r strftime(Sys.time(), "%Y-%m-%d, %H:%M %z")` in `r report_render_duration_in_seconds` seconds ( or `r report_render_duration_in_minutes` minutes)
+

manipulation/ellis-ua-admin.R

@@ -131,11 +131,11 @@ ds_map_hromada
 ds_map_settlement <- 
   #1
   ds_settlement %>% 
-  left_join(
+  inner_join(
     ds1 %>% distinct(settlement_code = level_4, hromada_code = level_3)
     ,by = "settlement_code"
   ) %>% 
-  left_join(
+  inner_join(
     ds_map_hromada
     ,by = "hromada_code"
   )
@@ -148,8 +148,8 @@ ds_map_hromada2 <-
   # select(-c("settlement_code","settlement_name", "settlement_type"))
   select(!starts_with("settlement_")) %>% 
   # View()
-  distinct() %>% 
-  filter(!is.na(hromada_name))
+  distinct()
+
 
 identical(ds_map_hromada, ds_map_hromada2)
 # Therefore we will use ds_map_settlement as the primary file