Example using crosswalk & data from other R packages?
Closed this issue · 3 comments
cynthiahqy commented
Consider calling crosswalks from countrycode or concordance and showing conversion to crossmap.
cynthiahqy commented
An example using comtradr, concordance and an early version of conformr -- though I recall there was an authentication requirement for comtrade data:
Demo dataset
- comtrade (2012-2007, for two countries, import)
library(comtradr)
q <- list()
q$Germany <- comtradr::ct_search(reporters = "Australia",
partners = c("Germany"),
start_date = 2016,
end_date = 2017,
type = "goods",
trade_direction = "imports",
commod_codes = c("All"))
q$Indonesia <- comtradr::ct_search(reporters = "Australia",
partners = c("Indonesia"),
start_date = 2016,
end_date = 2017,
type = "goods",
trade_direction = "imports",
commod_codes = "All")
monthly <- ct_search(reporters = "Australia",
partners = c("Indonesia"),
start_date = 2016,
end_date = 2016,
freq = "monthly",
type = "goods",
trade_direction = "imports",
commod_codes = "Total")
q$Indonesia %>% glimpse()
q$Indonesia %>%
filter(aggregate_level == 4, trade_flow == "Import") %>%
select(classification, period, reporter_iso, partner_iso, commodity_code, trade_value_usd)
concordance::hs5_hs4 %>%
select(ends_with("_4d")) %>%
distinct() %>%
conformr::make_cd_equal(., code_from = HS5_4d, code_to = HS4_4d) %>%
filter(n_dest != 1)
library(dplyr)
# TODO: highlight code, value 2D transformation.
make code-dict (equal weights)
## from codes
countrycode::<table> %>%
make_cd_equal(., code_A, code_B, direction = c("BA"))
## from code columns in data table
data %>%
distinct(code_A, code_B) %>%
make_cd_equal(., code_A, code_B, direction = c("AB")) %>%
check_cd_coverage(., data, code_from, code_to) ## should definitely pass
verify existing code-dict
code_dict %>%
check_cd_weights(., code_from, code_to, weight_col) %>%
check_cd_coverage(., data, code_from, code_to)
apply code-dict to data
data_A %>%
dplyr::group_by(case) %>%
conformr::convert(data = .,
code_dict = code_dict_BA,
code_from = code_A,
code_to = code_B,
values_from = value_A,
values_to = "value_B",
weight_col = weight)
cynthiahqy commented
First 6 codes of anzsco22 to isco8 crosswalk have unit & fractional weights.
cynthiahqy commented