The partners of trade are the variables in column each country that has entered into trade with the reporting country. There almost 200 countries in the world. But, each is not significant enough to be enlisted for its trade value. Therefore, some criteria needs to be set up for selecting the range of column variables. And, here we can choose GNI, the Gross National Income as the criterion for eliminating the countries whose trade is assumed to non-significant.
It is suitable for downloading the data for individual countries from the world bank database.
library(WDI)## Warning: package 'WDI' was built under R version 4.0.2
WDIsearch("gni.*current")## indicator name
## [1,] "NY.GNP.ATLS.CD" "GNI, Atlas method (current US$)"
## [2,] "NY.GNP.MKTP.CD" "GNI (current US$)"
## [3,] "NY.GNP.MKTP.CN" "GNI (current LCU)"
## [4,] "NY.GNP.MKTP.PC.CD" "GNI per capita (current US$)"
## [5,] "NY.GNP.MKTP.PP.CD" "GNI, PPP (current international $)"
## [6,] "NY.GNP.PCAP.CD" "GNI per capita, Atlas method (current US$)"
## [7,] "NY.GNP.PCAP.CN" "GNI per capita (current LCU)"
## [8,] "NY.GNP.PCAP.PP.CD" "GNI per capita, PPP (current international $)"
If you the name then “GNI (current US$)” is of relevance to us. We take its indicator which is “NY.GNP.MKTP.CD” to find and import the GNI figures for all the countries.
indicator <- "NY.GNP.MKTP.CD"
wdi_cat <- WDI(indicator = indicator, start = 2010, end = 2010, extra = T)
wdi_cat[1:10,]## iso2c country NY.GNP.MKTP.CD year
## 1 1A Arab World 2.091012e+12 2010
## 2 1W World 6.599366e+13 2010
## 3 4E East Asia & Pacific (excluding high income) 7.824143e+12 2010
## 4 7E Europe & Central Asia (excluding high income) 2.858405e+12 2010
## 5 8S South Asia 2.058046e+12 2010
## 6 AD Andorra NA 2010
## 7 AE United Arab Emirates 2.897035e+11 2010
## 8 AF Afghanistan 1.588451e+10 2010
## 9 AG Antigua and Barbuda 1.120980e+09 2010
## 10 AL Albania 1.180722e+10 2010
## iso3c region capital longitude latitude income lending
## 1 ARB NA NA
## 2 WLD NA NA
## 3 EAP NA NA
## 4 ECA NA NA
## 5 SAS NA NA
## 6 AND Z7 Andorra la Vella 1.5218 42.5075 XD XX
## 7 ARE ZQ Abu Dhabi 54.3705 24.4764 XD XX
## 8 AFG 8S Kabul 69.1761 34.5228 XM XI
## 9 ATG ZJ Saint John's -61.8456 17.1175 XD XF
## 10 ALB Z7 Tirane 19.8172 41.3317 XT XF
As you can see a sample of 10 rows where the country codes and GNI figures (“NY.GNP.MKTP.CD” column) are given is presented. And, we will be needing only the some of these columns for our data processing.
names(wdi_cat)## [1] "iso2c" "country" "NY.GNP.MKTP.CD" "year"
## [5] "iso3c" "region" "capital" "longitude"
## [9] "latitude" "income" "lending"
wdi2 <- wdi_cat %>%
select("country", "NY.GNP.MKTP.CD", "iso3c")
head(wdi2)## country NY.GNP.MKTP.CD iso3c
## 1 Arab World 2.091012e+12 ARB
## 2 World 6.599366e+13 WLD
## 3 East Asia & Pacific (excluding high income) 7.824143e+12 EAP
## 4 Europe & Central Asia (excluding high income) 2.858405e+12 ECA
## 5 South Asia 2.058046e+12 SAS
## 6 Andorra NA AND
So, there are 3 columns for our need, 1. country, 2. NY.GNP.MKTP.CD, 3. iso3c.
The last one is the country code that is necessary for joining the dataframe with our partners list we have.
Loading country_list from excel file downloaded from the UN Comtrade database.
https://unstats.un.org/unsd/tradekb/Knowledgebase/50377/Comtrade-Country-Code-and-Name
library(readxl)
country_list <- read_excel("comtradecountry_nowlist.xlsx")renaming the columns of the dataframe
colnames(country_list) <- c("country.code", "country.name", "iso3c")
head(country_list)## # A tibble: 6 x 3
## country.code country.name iso3c
## <dbl> <chr> <chr>
## 1 4 Afghanistan AFG
## 2 8 Albania ALB
## 3 12 Algeria DZA
## 4 16 American Samoa ASM
## 5 20 Andorra AND
## 6 24 Angola AGO
The WDI data is joined with the country_list so as to match the GNI for UN Comtrade list countries.
pp <- left_join(country_list, wdi2, by = "iso3c")
colnames(pp)[5] <- "GNI"
names(pp)## [1] "country.code" "country.name" "iso3c" "country" "GNI"
After joining the datasets the share of each country’s GNI against the world GNI is calculated. A new column share is created for enlisting the ratios of each country’s GNI.
pp$share <- round((pp$GNI/6.594123e+13)*100, 2)
head(pp)## # A tibble: 6 x 6
## country.code country.name iso3c country GNI share
## <dbl> <chr> <chr> <chr> <dbl> <dbl>
## 1 4 Afghanistan AFG Afghanistan 15884505938. 0.02
## 2 8 Albania ALB Albania 11807221086. 0.02
## 3 12 Algeria DZA Algeria 160820354235. 0.24
## 4 16 American Samoa ASM American Samoa NA NA
## 5 20 Andorra AND Andorra NA NA
## 6 24 Angola AGO Angola 75712633817. 0.11
removing NA rows by using complete.cases
pp1 <- na.omit(pp)
nrow(pp)## [1] 248
nrow(pp1)## [1] 191
As you can see the complete cases are only 191 out of total cases of 248. Now filtering only those countries whose GNI share is above 0.
pp2 <- pp1 %>%
filter(share > 0) %>%
select(country.code, country.name)
nrow(pp2)## [1] 156
The list of countries whose GNI is of some significance are 156 in total. So, for the panel data a list of 156 countries would be used for creating a block of matrix against the selected commodities. It would allow us only meaningful entities for datadownloading.
write.csv(pp2, "partner_list.csv", row.names = F)Above command shall export the list of 156 country.name and their country.code in a comma separated file (csv).