R/add_gwcode_to_cow.R
add_gwcode_to_cow.Rdadd_gwcode_to_cow() allows you to match, as well as one can, Gleditsch-Ward system membership data
with Correlates of War state system membership data.
add_gwcode_to_cow(data)
| data | a dyad-year data frame (either "directed" or "non-directed") or a state-year data frame. |
|---|
add_gwcode_to_cow() takes a dyad-year data frame or state-year data frame that already has Correlates of War
state system codes and adds their corollary Gleditsch-Ward codes.
The data-raw directory on the project's Github contains more information about the underlying data that assists
in merging in these codes.
The user will invariably need to be careful and ask why they want these data included. The issue here is that both have a different
composition and the merging process will not (and cannot) be perfect. We can note that a case like Serbia/Yugoslavia is not too difficult
to handle (since "Serbia" never overlaps with "Yugoslavia" in the Gleditsch-Ward data and Correlates of War understands Serbia as the
predecessor state, dominant state, and successor state to Yugoslavia). However, there is greater weirdness with a case like Yemen/Yemen
Arab Republic. The script will *not* create state-year or dyad-year duplicates for the Correlates of War codes. The size of the original data
remain unchanged. However, there will be some year duplicates for various Gleditsch-Ward codes (e.g. Yemen, again). Use with care. You can also use
the countrycode package. Whether you use this function or the countrycode package, do *not* do this kind of merging without assessing
the output.
Steven V. Miller
# just call `library(tidyverse)` at the top of the your script library(magrittr) cow_ddy %>% add_gwcode_to_cow()#> # A tibble: 2,063,670 x 5 #> ccode1 ccode2 year gwcode1 gwcode2 #> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 2 20 1920 2 20 #> 2 2 20 1921 2 20 #> 3 2 20 1922 2 20 #> 4 2 20 1923 2 20 #> 5 2 20 1924 2 20 #> 6 2 20 1925 2 20 #> 7 2 20 1926 2 20 #> 8 2 20 1927 2 20 #> 9 2 20 1928 2 20 #> 10 2 20 1929 2 20 #> # … with 2,063,660 more rows#>#> # A tibble: 16,731 x 4 #> ccode statenme year gwcode #> <dbl> <chr> <dbl> <dbl> #> 1 2 United States of America 1816 2 #> 2 2 United States of America 1817 2 #> 3 2 United States of America 1818 2 #> 4 2 United States of America 1819 2 #> 5 2 United States of America 1820 2 #> 6 2 United States of America 1821 2 #> 7 2 United States of America 1822 2 #> 8 2 United States of America 1823 2 #> 9 2 United States of America 1824 2 #> 10 2 United States of America 1825 2 #> # … with 16,721 more rows