add dhs script

[mcooper]

Hello,

Thanks for this fantastic package, I use it all the time.

One data source I frequently work with are the Demographic and Health Surveys, and they have their own unique country code format. It would be great to add their schema to the package.

Thanks!!

[cjyetman]

This is cool, thanks! The build failed, but it looks like it was for a completely unrelated reason... Error: package ‘R6’ was installed before R 4.0.0: please re-install it

[cjyetman]

In case anyone wonders, there is some significant difference between this and iso2c...

library(jsonlite)
library(dplyr)
library(countrycode)
url <- 'https://api.dhsprogram.com/rest/dhs/countries?select=DHS_CountryCode,CountryName&f=json'
fromJSON(url)[['Data']] %>%
  dplyr::select(CountryName, DHS_CountryCode) %>% 
  mutate(iso2c = countrycode(CountryName, 'country.name', 'iso2c')) %>% 
  filter(DHS_CountryCode != iso2c)

#>             CountryName DHS_CountryCode iso2c
#> 1              Botswana              BT    BW
#> 2               Burundi              BU    BI
#> 3    Dominican Republic              DR    DO
#> 4           El Salvador              ES    SV
#> 5     Equatorial Guinea              EK    GQ
#> 6             Guatemala              GU    GT
#> 7                 India              IA    IN
#> 8            Kazakhstan              KK    KZ
#> 9       Kyrgyz Republic              KY    KG
#> 10              Liberia              LB    LR
#> 11           Madagascar              MD    MG
#> 12              Moldova              MB    MD
#> 13              Namibia              NM    NA
#> 14            Nicaragua              NC    NI
#> 15                Niger              NI    NE
#> 16 Nigeria (Ondo State)              OS    NG

[mcooper]

Just FYI after seeing @cjyetman print the first 16 lines, I realized there is a duplicate in there. There was one instance of a survey conducted in Ondo state Nigeria, so that survey had its own code. I filtered out that one, and checked that all the rest are at the country level and there are no duplicates.

[vincentarelbundock]

Thanks a lot @mcooper !

I merged the script and rebuilt the dictionary. If you could please try the Github Master version, I could then upload to CRAN:

remotes::install_github('vincentarelbundock/countrycode')
library(countrycode)
countrycode(c('Afghanistan', 'Namibia', 'Liberia'), 'country.name', 'dhs')

[vincentarelbundock]

Thanks @cjyetman for comparing the two sets. Very useful!

[mcooper]

Just tried it locally on a project I'm in the middle of, it works great. Thanks for updating the package so quickly!

Mat

On Thu, May 7, 2020 at 1:53 PM Vincent Arel-Bundock < notifications@github.com> wrote:

Thanks @cjyetman https://github.com/cjyetman for comparing the two sets. Very useful!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/vincentarelbundock/countrycode/pull/218#issuecomment-625405728, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEAGRK7EC2HWOELE47DXHS3RQLYTDANCNFSM4M3PYLZA .

[vincentarelbundock]

This is much easier when people submit actual working code ;)

[vincentarelbundock]