Creating gravity datasets

knitr::opts_chunk$set(eval = FALSE)


Gravity dataset with zero trade flows is an edited version of the full gravity dataset that is used in @Head2010.

Dataset gravity_no_zeros corresponds to the dataset without zero trade flows, gravity_zeros, on the other hand, includes zero trade flows.


In order to have a dataset suited for all functions, a cross-sectional dataset is chosen. All incomplete rows and observations with missing trade flows are therefore excluded from the dataset.

As some of the functions in the package are capable of handling zero values in trade flows and some are not, two datasets, gravity_zeros and gravity_no_zeros, are provided.

The original dataset was downloaded from SciencesPo but the original link is not available anymore. It was edited in the following way:

# 1: Import and read the dataset

url <- ""
zip <- ""

if (!file.exists(zip)) { try(download.file(url, zip)) }
try(system("7z e -aos"))

col_regfile09 <- read_dta("col_regfile09.dta")

# 2: Isolation of one year

data06 <- col_regfile09 %>% 
  filter(year == 2006)

# 3: Choosing variables

data06 <- data06 %>% 
  select(iso_o, iso_d, distw, gdp_o, gdp_d, rta, flow, contig, comlang_off, comcur)

# 4: Isolation of complete cases

gravity_zeros <- data06 %>% 

# 5: Exclusion of trade flows equal to 0

gravity_no_zeros <- gravity_zeros %>% 
  filter(flow > 0)

# 6: Export the data

save(gravity_zeros, file = "gravity_zeros.rdata", compress = "xz")
save(gravity_no_zeros, file = "gravity_no_zeros.rdata", compress = "xz")


Try the gravity package in your browser

Any scripts or data that you put into this service are public.

gravity documentation built on July 27, 2020, 9:08 a.m.