In annecori/mRIIDSprocessData: Epidemiological forcasting using real-time data

library(magrittr)
library(ggplot2)
library(dplyr)

In the WHO dataset, location of residence was recorded to the second administrative level, termed prefecture in Guinea, county in Liberia and district in Sierra Leone. There are minor differences between the names recorded in the WHO dataset and as available in the dataset recording their latitude, longitude and population. We will clean up these differences and save the clean files.

infile <- here::here("data/CaseCounts/processed/", "WHO_bydistricts.csv")
WHO_districts <- readr::read_csv(infile) 

guinea_sierra <- here::here("data","Geography/GravityModel/raw/adm2-fixed.txt") %>%
                     read.csv(stringsAsFactors = FALSE,
                              sep = "\t",
                              header = TRUE) %>%
                    filter(ADM0 %in% c("Guinea", "Sierra Leone"))  

liberia <- here::here("data", "Geography/GravityModel/raw/adm1_centroids_fixed.tsv") %>%
                   read.csv(stringsAsFactors = FALSE,
                            sep = "\t",
                            header = TRUE) %>%
                   filter(ADM0 == "Liberia")  

guinea_sierra %<>%
    select(ADM0, ADM2, Centroid_Lon, Centroid_Lat, Pop) %<>%
    rename(CL_DistrictRes = ADM2)

liberia %<>%
    select(ADM0, ADM1, Centroid_Lon, Centroid_Lat, Pop) %<>%
    rename(CL_DistrictRes = ADM1)


all_centroids <- bind_rows(guinea_sierra, liberia) 

We now convert all non-ASCII letters to ASCII characters (e.g., accented e is converted to e), remove spaces and convert case.

all_centroids$CL_DistrictRes     %<>%
    iconv(to='ASCII//TRANSLIT')  %<>%
    gsub("\\'", '', .)  %<>%
    gsub(' ', '', .) %<>%
    toupper %<>%
    factor

all_centroids$ADM0   %<>%
    gsub(' ', '', .) %<>%
    toupper 

In the WHO data-set, no distinction has been made between Western Urban and Western Rural districts of Sierra Leone. So we treat the co-ordinates of Western Urban as those of Western district.

all_centroids$CL_DistrictRes %<>% gsub("WESTERNURBAN", "WESTERN", .)

Some of the prefectures of Guinea are not represented in the WHO data. These are: Gaoual, Koubia, Koundara, Labe, Lelouma, Mamou and Mandiana. Hence we will drop these from the data frame of centroids.

missing <- setdiff(unique(all_centroids$CL_DistrictRes),
                   unique(WHO_districts$CL_DistrictRes))

notmissing    <- !(all_centroids$CL_DistrictRes %in% missing)
all_centroids <- all_centroids[notmissing, ]


all_centroids$CL_DistrictRes %<>% factor

all_centroids$CL_DistrictRes %<>%
    plyr::mapvalues(from = c("KISSIDOUGO",  "YOMOU",  "N'ZEREKORE", "GBARPOLU" ),
                    to   = c("KISSIDOUGOU", "YAMOU",  "NZEREKORE", "GBAPOLU"))

Both files are now consistent and we can write them out.

outfile <- here::here("data/Geography/GravityModel/processed/", "all_centroids.csv")
readr::write_csv(all_centroids, path = outfile)

annecori/mRIIDSprocessData documentation built on May 29, 2019, 1:16 p.m.