In stix-global/eutradeflows: Download bilateral trade flow data from Eurostat Comext

knitr::opts_chunk$set(echo = TRUE)
library(networkD3)
library(dplyr)
# library(tidyverse)

Notes

Future work

This sankey diagram can be integrated in a It could preload data for all years and have a slider for year. If the data is not too large, the slider could

SHiny network example

Adding a slider input with a play button

See also

• original development of the Sankey diagram in the Statistical learning RStudio project folder.
• It's possible to use the Sankey diagram in a shiny interface.

Errors

Reporter and partner should have different codes.

Error: It looks like Source/Target is not zero-indexed. This is required in JavaScript and so your plot may not render. The problem is that countries in the nodes data frame are only indexed by position (javascript tables are zero-indexed) there is no code variable in the nodes table

Gerating sankey diagram for one product

library(eutradeflows)
con <- RMariaDB::dbConnect(RMariaDB::MariaDB(), dbname = "tradeflows")
# Raw Eurostat Comext data
wood <- tbl(con, "vld_comext_monthly") %>% 
    filter(productcode ==  44072199 & 
               flowcode == 1 & 
               period == "201702") %>% 
    # head(200) %>% 
    collect()

reporter <- tbl(con, "vld_comext_reporter") %>% collect()
partner <- tbl(con, "vld_comext_partner") %>% collect() 

wood2 <- wood %>%
    # Remove partner 1010 and reporter 0
    filter(partnercode != 1010 & partnercode != 1011 & 
               reportercode !=0) %>% 
    preparesankeynodes(reporter, partner, debugname = TRUE)

wood2 %>% plotsankey(value = "tradevalue", units = "K€")
wood2 %>% plotsankey(value = "quantity", units = "M3")

Sankey diagram for aggregated products

wood <- tbl(con, "vld_comext_monthly") %>% 
    filter(productcode %like%  '4407%' & 
               flowcode == 1 & 
               period == "201705") %>% 
    group_by(reportercode, partnercode) %>% 
    summarise(tradevalue = sum(tradevalue)) %>% 
    collect()
summary(wood$tradevalue)

# Largest sawnwood tradeflows
wood %>% 
    filter(partnercode != 1010 & partnercode != 1011 & 
               reportercode !=0) %>% 
    filter(tradevalue > 2000) %>%
    preparesankeynodes(reporter, partner) %>% 
    plotsankey()

January - August 2017 Sawnwood trade between VPA countries and Europe

Cameroon, Central African Republic, Ghana, Indonesia, Liberia, Republic of the Congo

Negotiating Côte d'Ivoire, Democratic Republic of the Congo, Gabon, Guyana, Honduras, Laos, Malaysia, Thailand, Vietnam Contact

vpaimplementing <- c(302,306,276,700,268,318)
vpanegotiating <- c(272,322,314,488,424,684,701,680,690)
vpa <- c(vpaimplementing, vpanegotiating)



# Sawnwood tradeflows with VPA countries
wood <- tbl(con, "vld_comext_monthly") %>% 
    filter(productcode %like%  '4407%' & 
               flowcode == 1 & 
               period >= 201701) %>% 
    group_by(reportercode, partnercode) %>% 
    summarise(tradevalue = sum(tradevalue)) %>% 
    collect()
summary(wood$tradevalue)


wood %>% 
    filter(partnercode != 1010 & partnercode != 1011 & 
               reportercode !=0) %>% 
    filter(partnercode %in% vpa) %>%
    preparesankeynodes(reporter, partner) %>% 
    plotsankey()

Plywood trade between VPA countries and the EU

wood <- tbl(con, "vld_comext_monthly") %>% 
    filter(productcode %like%  '4412%' & 
               flowcode == 1 & 
               period >= 201701) %>% 
    group_by(reportercode, partnercode) %>% 
    summarise(tradevalue = sum(tradevalue)) %>% 
    collect()
summary(wood$tradevalue)


wood %>% 
    filter(partnercode != 1010 & partnercode != 1011 & 
               reportercode !=0) %>% 
    filter(partnercode %in% vpa) %>%
    preparesankeynodes(reporter, partner) %>% 
    plotsankey()

Furniture trade between VPA countries and the EU

940330 Furniture; wooden, for office use 940340 Furniture; wooden, for kitchen use 940350 Furniture; wooden, for bedroom use 940360 Furniture; wooden, other than for office, kitchen or bedroom u

wood <- tbl(con, "vld_comext_monthly") %>% 
    filter(productcode %like%  '94033%' | 
               productcode %like%  '94034%' |
               productcode %like%  '94035%' |
               productcode %like%  '94036%') %>% 
    filter(flowcode == 1 & period >= 201701) %>% 
    group_by(reportercode, partnercode) %>% 
    summarise(tradevalue = sum(tradevalue)) %>% 
    collect()
summary(wood$tradevalue)


wood %>% 
    filter(partnercode != 1010 & partnercode != 1011 & 
               reportercode !=0) %>% 
    filter(partnercode %in% vpa) %>%
    preparesankeynodes(reporter, partner) %>% 
    plotsankey()

Sankey diagram for many products

Loop per month

Loop per product for the sum(2017) data

stix-global/eutradeflows documentation built on Nov. 13, 2020, 9:23 p.m.