Clean, find facilities to code

scratch.R

@@ -0,0 +1,40 @@
+# Scratch file
+library(janitor)
+library(dplyr)
+library(readr)
+Geo2017 <- clean_names(read_csv("data/AllDHBsGeocoded.csv"), case = "snake") %>%
+    select(-facility_opening_date) %>%
+    mutate(dhb_name = ifelse(is.na(dhb_name), "Missing", dhb_name),
+           health_facility_code = ifelse(is.na(health_facility_code),
+                                         "Missing", health_facility_code))
+    
+library(readxl)
+newFacil <- clean_names(read_excel("data/facilities20190902.xlsx"), 
+                        case = "snake") %>%
+    filter(facility_type_name == "GP Practice") %>% 
+    select(-facility_closing_date, -facility_opening_date) %>%
+    mutate(dhb_name = ifelse(is.na(dhb_name), "Missing", dhb_name),
+           health_facility_code = ifelse(is.na(health_facility_code),
+                                         "Missing", health_facility_code))
+    
+DHBs <- unique(c(newFacil$dhb_name, Geo2017$dhb_name))
+unique(newFacil$dhb_name) %in% unique(Geo2017$dhb_name)
+
+DHBs[!(DHBs %in% newFacil$dhb_name)]
+DHBs[!(DHBs %in% Geo2017$dhb_name)]
+nrow(newFacil)
+
+nrow(Geo2017)
+
+colSums(is.na(newFacil))
+colSums(is.na(Geo2017))
+
+gp_prac <- left_join(newFacil, Geo2017, by = c("name", "health_facility_code", 
+                                               "hpi_facility_id",
+                                               "address", "dhb_name",
+                                               "fac_type", "facility_type_name"))
+colSums(is.na(gp_prac))
+
+gp_same <- filter(gp_prac, !is.na(gp_prac$calculated_address))
+gp_new <- filter(gp_prac, is.na(gp_prac$calculated_address)) %>%
+    select(-lon, -lat, -calculated_address)