The structure of BACI dataset is described here.
I downloaded the BACI data (2.2G ZIP) from CEPII and extracted the CSV file for the year 2000 for experimenting.
> wdir <- "C:/Users/vlado/docs/papers/2023/HiTeC/London/slides/BACI/95-21"
> setwd(wdir)
> source("https://raw.githubusercontent.com/bavla/Rnet/master/R/Pajek.R")
>
> C <- read.csv("country_codes_V202301.csv")
> head(C)
country_code country_name_abbreviation country_name_full iso_2digit_alpha iso_3digit_alpha
1 4 Afghanistan Afghanistan AF AFG
2 8 Albania Albania AL ALB
3 12 Algeria Algeria DZ DZA
4 16 American Samoa American Samoa AS ASM
5 20 Andorra Andorra AD AND
6 24 Angola Angola AO AGO
> B <- read.csv("BACI_HS92_Y2000_V202301.csv")
> str(B)
'data.frame': 7271864 obs. of 6 variables:
$ t: int 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 ...
$ i: int 4 4 4 4 4 4 4 4 4 4 ...
$ j: int 12 12 12 12 12 12 20 20 20 20 ...
$ k: int 90920 90930 90940 121190 330300 848330 300490 570110 570210 570242 ...
$ v: num 2.823 11.216 7.981 2.051 0.552 ...
$ q: chr " 6.800" " 11.550" " 19.140" " 2.000" ...
> N <- aggregate(B$v,by=list(B$i,B$j),FUN=sum)
> str(N)
'data.frame': 26043 obs. of 3 variables:
$ Group.1: int 8 31 36 40 50 56 76 100 112 124 ...
$ Group.2: int 4 4 4 4 4 4 4 4 4 4 ...
$ x : num 0.209 757.165 61.238 506.723 4008.518 ...
> sum(B$v[1:6])
[1] 24.812
> N[which((N$Group.1==4)&(N$Group.2==12)),]
Group.1 Group.2 x
178 4 12 24.812
> colnames(N) <- c("u","v","w")
> cc <- C$country_code
> N$u <- as.integer(factor(N$u,level=cc))
> N$v <- as.integer(factor(N$v,level=cc))
> uvLab2net(C$iso_2digit_alpha,N$u,N$v,N$w,rep(2000,nrow(N)),Net="WT00.net",dir=TRUE)
On the graphical representation in Pajek, I noticed that some country labels (DEU, BEL, …) appear twice. I checked them in source data.
> which(C$iso_3digit_alpha=="DEU") [1] 79 81 > C[c(79,81),] country_code country_name_abbreviation country_name_full iso_2digit_alpha iso_3digit_alpha 79 276 Germany Germany DE DEU 81 280 Fmr Fed. Rep. of Germany Former Federal Republic of Germany DE DEU
> zdir <- unzip("../BACI_HS92_V202301.zip",list=TRUE)
> zdir
Name Length Date
1 BACI_HS92_Y1995_V202301.csv 238112541 2023-01-31 11:11:00
2 BACI_HS92_Y1996_V202301.csv 255110534 2023-01-31 11:13:00
3 BACI_HS92_Y1997_V202301.csv 268464528 2023-01-31 11:14:00
4 BACI_HS92_Y1998_V202301.csv 279524363 2023-01-31 11:16:00
5 BACI_HS92_Y1999_V202301.csv 286562400 2023-01-31 11:18:00
6 BACI_HS92_Y2000_V202301.csv 347577043 2023-01-31 11:20:00
7 BACI_HS92_Y2001_V202301.csv 359134170 2023-01-31 11:22:00
8 BACI_HS92_Y2002_V202301.csv 369188019 2023-01-31 11:24:00
9 BACI_HS92_Y2003_V202301.csv 381371385 2023-01-31 11:27:00
10 BACI_HS92_Y2004_V202301.csv 398264263 2023-01-31 11:29:00
11 BACI_HS92_Y2005_V202301.csv 412270441 2023-01-31 11:32:00
12 BACI_HS92_Y2006_V202301.csv 428661613 2023-01-31 11:34:00
13 BACI_HS92_Y2007_V202301.csv 439195266 2023-01-31 11:37:00
14 BACI_HS92_Y2008_V202301.csv 445018106 2023-01-31 11:40:00
15 BACI_HS92_Y2009_V202301.csv 440450004 2023-01-31 11:42:00
16 BACI_HS92_Y2010_V202301.csv 451977993 2023-01-31 11:45:00
17 BACI_HS92_Y2011_V202301.csv 459970011 2023-01-31 11:48:00
18 BACI_HS92_Y2012_V202301.csv 473158279 2023-01-31 11:51:00
19 BACI_HS92_Y2013_V202301.csv 482756570 2023-01-31 11:54:00
20 BACI_HS92_Y2014_V202301.csv 483932125 2023-01-31 11:57:00
21 BACI_HS92_Y2015_V202301.csv 500816869 2023-01-31 12:00:00
22 BACI_HS92_Y2016_V202301.csv 500809051 2023-01-31 12:03:00
23 BACI_HS92_Y2017_V202301.csv 511566990 2023-01-31 12:06:00
24 BACI_HS92_Y2018_V202301.csv 515154942 2023-01-31 12:09:00
25 BACI_HS92_Y2019_V202301.csv 517328358 2023-01-31 12:13:00
26 BACI_HS92_Y2020_V202301.csv 499104038 2023-01-31 12:15:00
27 BACI_HS92_Y2021_V202301.csv 509386838 2023-01-31 12:22:00
28 country_codes_V202301.csv 10431 2023-02-03 11:11:00
29 product_codes_HS92_V202301.csv 543275 2023-02-03 13:58:00
> source("https://raw.githubusercontent.com/bavla/Rnet/master/R/Pajek.R")
> C <- read.csv("country_codes_V202301.csv")
> cc <- C$country_code; cn <- C$iso_2digit_alpha; cn[139] <- "NA"
> zdir <- unzip("../BACI_HS92_V202301.zip",list=TRUE)
> for(i in 1:27){
+ year <- 1994+i; cat(year,":",date(),"\n"); flush.console()
+ net <- paste("WT",year,".net",sep="")
+ B <- read.csv(unzip("../BACI_HS92_V202301.zip",files=zdir$Name[i]))
+ file.remove(zdir$Name[i])
+ N <- aggregate(B$v,by=list(B$i,B$j),FUN=sum)
+ colnames(N) <- c("u","v","w")
+ N$u <- as.integer(factor(N$u,level=cc))
+ N$v <- as.integer(factor(N$v,level=cc))
+ uvLab2net(cn,N$u,N$v,N$w,rep(year,nrow(N)),Net=net,dir=TRUE)
+ }
> cat("****",":",date(),"\n")
1995 : Fri Aug 11 01:59:58 2023
1996 : Fri Aug 11 02:00:12 2023
1997 : Fri Aug 11 02:00:27 2023
...
2020 : Fri Aug 11 02:14:01 2023
2021 : Fri Aug 11 02:14:44 2023
**** : Fri Aug 11 02:15:29 2023
> source("https://raw.githubusercontent.com/bavla/Rnet/master/R/Pajek.R")
> C <- read.csv("country_codes_V202301.csv")
> cc <- C$country_code; cn <- C$iso_2digit_alpha; cn[139] <- "NA"
> net <- file("WTtime.net","w",encoding="UTF-8")
> act <- file("activity.txt","w",encoding="UTF-8")
> n <- length(cn); tin <- '" [1995-2021]\n'
> cat("% WorldTrade",date(),"\n*vertices",n,"\n",file=net)
> for(v in 1:n) cat(v,' "',cn[v],tin,sep="",file=net)
> cat("*arcs\n",file=net)
> zdir <- unzip("../BACI_HS92_V202301.zip",list=TRUE)
> for(i in 1:27){
> year <- 1994+i; cat(year,":",date(),"\n"); flush.console()
> B <- read.csv(unzip("../BACI_HS92_V202301.zip",files=zdir$Name[i]))
> file.remove(zdir$Name[i])
> N <- aggregate(B$v,by=list(B$i,B$j),FUN=sum)
> colnames(N) <- c("u","v","w")
> N$u <- as.integer(factor(N$u,level=cc))
> N$v <- as.integer(factor(N$v,level=cc))
> for(e in 1:nrow(N)) cat(N$u[e],N$v[e],N$w[e]," [",year,"]\n",file=net)
> cat(year,setdiff(1:n,union(N$u,N$v)),"\n",file=act)
> }
> cat("****",":",date(),"\n"); close(act); close(net)
1995 : Fri Aug 11 03:12:59 2023
1996 : Fri Aug 11 03:13:13 2023
1997 : Fri Aug 11 03:13:28 2023
...
2019 : Fri Aug 11 03:23:15 2023
2020 : Fri Aug 11 03:23:47 2023
2021 : Fri Aug 11 03:24:19 2023
**** : Fri Aug 11 03:24:51 2023
For each year we also determined the nonactive countries
1995 4 17 23 48 56 78 80 81 88 115 120 134 139 144 146 147 176 183 187 190 197 201 202 205 224 1996 4 17 23 48 56 78 80 81 88 115 120 134 139 144 146 147 176 183 187 197 201 202 205 224 1997 4 17 23 48 56 78 80 81 88 115 120 134 139 144 146 147 176 183 187 190 197 201 202 205 224 1998 4 17 23 48 56 78 80 81 88 115 120 134 139 144 146 147 176 183 187 197 201 202 205 224 1999 4 18 23 48 56 78 80 81 88 115 134 139 144 146 147 176 183 187 190 197 201 202 205 224 2000 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 2001 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 2002 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 2003 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 229 2004 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 229 2005 18 56 80 81 134 144 146 147 176 187 190 198 201 202 224 229 2006 18 56 80 81 144 146 147 176 190 198 201 202 224 229 237 2007 18 56 80 81 144 146 147 176 190 198 201 202 224 229 237 2008 18 56 80 81 144 146 147 176 190 198 201 202 224 229 237 2009 18 56 80 81 144 146 147 176 190 198 201 202 224 229 237 2010 18 56 80 81 146 147 176 190 198 201 202 224 229 237 2011 18 56 80 81 176 190 198 201 202 224 229 237 2012 18 56 80 81 143 176 190 198 203 224 229 237 2013 18 56 80 81 143 190 198 203 224 229 237 2014 18 48 56 80 81 143 190 198 203 224 229 237 2015 18 48 56 80 81 143 190 198 203 224 229 237 2016 18 48 56 80 81 143 190 198 203 224 229 237 2017 18 48 56 80 81 143 190 198 203 224 229 237 2018 18 48 56 80 81 190 198 203 224 229 237 2019 18 48 56 80 81 143 190 198 203 224 229 237 2020 18 48 56 80 81 143 190 198 203 224 229 237 2021 18 48 56 80 81 143 190 198 203 224 229 237
> library(jsonlite)
> sec <- rep(0,99)
> sec[1:5] <- 1; sec[6:14] <- 2; sec[15] <- 3; sec[16:24] <- 4
> sec[25:27] <- 5; sec[28:38] <- 6; sec[39:40] <- 7; sec[41:43] <- 8
> sec[44:46] <- 9; sec[47:49] <- 10; sec[50:63] <- 11; sec[64:67] <- 12
> sec[68:70] <- 13; sec[71] <- 14; sec[72:83] <- 15; sec[84:85] <- 16
> sec[86:89] <- 17; sec[90:92] <- 18; sec[93] <- 19; sec[94:96] <- 20
> sec[97:98] <- 21; sec[99] <- 22
> C <- read.csv("country_codes_V202301.csv")
> cc <- C$country_code; cn <- C$iso_2digit_alpha; cn[139] <- "NA"
> c3 <- C$iso_3digit_alpha; ca <- C$country_name_abbreviation
> S <- read.csv("../sectShort.txt",sep="",head=TRUE)
> Ch <- read.csv("../chapShort.txt",sep=";",head=TRUE)
> n <- length(cn)
> info <- list(network="WTchapters",
+ title="BACI/CEPII World trade flows 1995-2021",
+ by="Gaulier, G. and Zignago, S.",
+ ref="Gaulier, G. and Zignago, S. (2010) BACI: International Trade Database at the Product-Level.
The 1994-2007 Version. CEPII Working Paper, N°2010-23.",
+ href="http://www.cepii.fr/CEPII/en/bdd_modele/bdd_modele_item.asp?id=37",
+ creator="V. Batagelj",
+ date=date() )
> ways <- list(ex="exporter",im="importer",ch="chapter",y="year")
> nodes <- list(ex=data.frame(ID=cn,iso3=c3,long=ca),im=data.frame(ID=cn),
+ ch=data.frame(ID=trimws(Ch$lab)),y=data.frame(ID=Y))
> data <- list(sectID=S$lab,sect=sec)
> zdir <- unzip("../BACI_HS92_V202301.zip",list=TRUE)
> V <- NULL
> for(i in 1:27){
+ year <- 1994+i; cat(year,":",date(),"\n"); flush.console()
+ B <- read.csv(unzip("../BACI_HS92_V202301.zip",files=zdir$Name[i]))
+ file.remove(zdir$Name[i]); B$k <- as.integer(B$k %/% 10000)
+ N <- aggregate(B$v,by=list(B$i,B$j,B$k,B$t),FUN=sum)
+ colnames(N) <- c("ex","im","ch","y","w")
+ N$ex <- as.integer(factor(N$ex,level=cc))
+ N$im <- as.integer(factor(N$im,level=cc))
+ N$y <- as.integer(factor(N$y,level=1995:2021))
+ V <- rbind(V,N)
+ }
> cat("****",":",date(),"\n")
> MN <- list(format="MWnets",info=info,ways=ways,nodes=nodes,links=V,data=data)
> write(toJSON(MN),"WTchap.json")
1995 : Sat Aug 12 00:47:24 2023
1996 : Sat Aug 12 00:47:45 2023
1997 : Sat Aug 12 00:48:08 2023
1998 : Sat Aug 12 00:48:33 2023
...
2018 : Sat Aug 12 01:04:08 2023
2019 : Sat Aug 12 01:05:14 2023
2020 : Sat Aug 12 01:06:17 2023
2021 : Sat Aug 12 01:07:22 2023
> str(MN)
List of 6
$ format: chr "MWnets"
$ info :List of 7
..$ network: chr "WTchapters"
..$ title : chr "BACI/CEPII World trade flows 1995-2021"
..$ by : chr "Gaulier, G. and Zignago, S."
..$ ref : chr "Gaulier, G. and Zignago, S. (2010) BACI: International Trade Database at the Product-Level.
The 1994-2007 Versi"| __truncated__
..$ href : chr "http://www.cepii.fr/CEPII/en/bdd_modele/bdd_modele_item.asp?id=37"
..$ creator: chr "V. Batagelj"
..$ date : chr "Sat Aug 12 00:47:23 2023"
$ ways :List of 4
..$ ex: chr "exporter"
..$ im: chr "importer"
..$ ch: chr "chapter"
..$ y : chr "year"
$ nodes :List of 4
..$ ex:'data.frame': 238 obs. of 3 variables:
.. ..$ ID : chr [1:238] "AF" "AL" "DZ" "AS" ...
.. ..$ iso3: chr [1:238] "AFG" "ALB" "DZA" "ASM" ...
.. ..$ long: chr [1:238] "Afghanistan" "Albania" "Algeria" "American Samoa" ...
..$ im:'data.frame': 238 obs. of 1 variable:
.. ..$ ID: chr [1:238] "AF" "AL" "DZ" "AS" ...
..$ ch:'data.frame': 99 obs. of 1 variable:
.. ..$ ID: chr [1:99] "Live Animals and poultry" "Animal Meat and edible offals" "Fish, fillets and aquatic meat" "Honey, Eggs and Dairy" ...
..$ y :'data.frame': 27 obs. of 1 variable:
.. ..$ ID: chr [1:27] "1995" "1996" "1997" "1998" ...
$ links :'data.frame': 20384334 obs. of 5 variables:
..$ ex: int [1:20384334] 71 85 95 102 142 223 18 71 79 95 ...
..$ im: int [1:20384334] 2 2 2 2 2 2 3 3 3 3 ...
..$ ch: int [1:20384334] 1 1 1 1 1 1 1 1 1 1 ...
..$ y : int [1:20384334] 1 1 1 1 1 1 1 1 1 1 ...
..$ w : num [1:20384334] 352.82 126.256 225 398.26 0.997 ...
$ data :List of 2
..$ sectID: chr [1:21] "animals" "vegetables" "fats&oils" "foodstuffs" ...
..$ sect : num [1:99] 1 1 1 1 1 2 2 2 2 2 ...
>
HS 2002 Classification by Section
> library(jsonlite)
> sec <- rep(0,99)
> sec[1:5] <- 1; sec[6:14] <- 2; sec[15] <- 3; sec[16:24] <- 4
> sec[25:27] <- 5; sec[28:38] <- 6; sec[39:40] <- 7; sec[41:43] <- 8
> sec[44:46] <- 9; sec[47:49] <- 10; sec[50:63] <- 11; sec[64:67] <- 12
> sec[68:70] <- 13; sec[71] <- 14; sec[72:83] <- 15; sec[84:85] <- 16
> sec[86:89] <- 17; sec[90:92] <- 18; sec[93] <- 19; sec[94:96] <- 20
> sec[97:98] <- 21; sec[99] <- 22
> C <- read.csv("country_codes_V202301.csv")
> cc <- C$country_code; cn <- C$iso_2digit_alpha; cn[139] <- "NA"
> c3 <- C$iso_3digit_alpha; ca <- C$country_name_abbreviation
> S <- read.csv("../sectShort.txt",sep="",head=TRUE)
> Ch <- read.csv("../chapShort.txt",sep=";",head=TRUE)
> n <- length(cn)
> info <- list(network="WTsections",
+ title="BACI/CEPII World trade flows 1995-2021",
+ by="Gaulier, G. and Zignago, S.",
+ ref="Gaulier, G. and Zignago, S. (2010) BACI: International Trade Database at the Product-Level.
The 1994-2007 Version. CEPII Working Paper, N°2010-23.",
+ href="http://www.cepii.fr/CEPII/en/bdd_modele/bdd_modele_item.asp?id=37",
+ creator="V. Batagelj",
+ date=date() )
> ways <- list(ex="exporter",im="importer",s="section",y="year")
> nodes <- list(ex=data.frame(ID=cn,iso3=c3,long=ca),im=data.frame(ID=cn),
+ s=data.frame(ID=S$lab),y=data.frame(ID=Y))
> data <- list()
> zdir <- unzip("../BACI_HS92_V202301.zip",list=TRUE)
> V <- NULL
> for(i in 1:27){
+ year <- 1994+i; cat(year,":",date(),"\n"); flush.console()
+ B <- read.csv(unzip("../BACI_HS92_V202301.zip",files=zdir$Name[i]))
+ file.remove(zdir$Name[i]); B$k <- sec[as.integer(B$k %/% 10000)]
+ N <- aggregate(B$v,by=list(B$i,B$j,B$k,B$t),FUN=sum)
+ colnames(N) <- c("ex","im","s","y","w")
+ N$ex <- as.integer(factor(N$ex,level=cc))
+ N$im <- as.integer(factor(N$im,level=cc))
+ N$y <- as.integer(factor(N$y,level=1995:2021))
+ V <- rbind(V,N)
+ }
> cat("****",":",date(),"\n")
1995 : Sat Aug 12 01:45:32 2023
1996 : Sat Aug 12 01:45:50 2023
1997 : Sat Aug 12 01:46:10 2023
1998 : Sat Aug 12 01:46:33 2023
...
2018 : Sat Aug 12 02:02:17 2023
2019 : Sat Aug 12 02:03:15 2023
2020 : Sat Aug 12 02:04:13 2023
2021 : Sat Aug 12 02:05:06 2023
**** : Sat Aug 12 02:06:03 2023
> MN <- list(format="MWnets",info=info,ways=ways,nodes=nodes,links=V,data=data)
> write(toJSON(MN),"WTsect.json")
> str(MN)
List of 6
$ format: chr "MWnets"
$ info :List of 7
..$ network: chr "WTsections"
..$ title : chr "BACI/CEPII World trade flows 1995-2021"
..$ by : chr "Gaulier, G. and Zignago, S."
..$ ref : chr "Gaulier, G. and Zignago, S. (2010) BACI: International Trade Database at the Product-Level.
The 1994-2007 Versi"| __truncated__
..$ href : chr "http://www.cepii.fr/CEPII/en/bdd_modele/bdd_modele_item.asp?id=37"
..$ creator: chr "V. Batagelj"
..$ date : chr "Sat Aug 12 01:45:32 2023"
$ ways :List of 4
..$ ex: chr "exporter"
..$ im: chr "importer"
..$ s : chr "section"
..$ y : chr "year"
$ nodes :List of 4
..$ ex:'data.frame': 238 obs. of 3 variables:
.. ..$ ID : chr [1:238] "AF" "AL" "DZ" "AS" ...
.. ..$ iso3: chr [1:238] "AFG" "ALB" "DZA" "ASM" ...
.. ..$ long: chr [1:238] "Afghanistan" "Albania" "Algeria" "American Samoa" ...
..$ im:'data.frame': 238 obs. of 1 variable:
.. ..$ ID: chr [1:238] "AF" "AL" "DZ" "AS" ...
..$ s :'data.frame': 21 obs. of 1 variable:
.. ..$ ID: chr [1:21] "animals" "vegetables" "fats&oils" "foodstuffs" ...
..$ y :'data.frame': 27 obs. of 1 variable:
.. ..$ ID: chr [1:27] "1995" "1996" "1997" "1998" ...
$ links :'data.frame': 7575645 obs. of 5 variables:
..$ ex: int [1:7575645] 43 53 77 100 112 142 210 228 10 11 ...
..$ im: int [1:7575645] 1 1 1 1 1 1 1 1 2 2 ...
..$ s : num [1:7575645] 1 1 1 1 1 1 1 1 1 1 ...
..$ y : int [1:7575645] 1 1 1 1 1 1 1 1 1 1 ...
..$ w : num [1:7575645] 84 6.99 21.07 303.3 849.5 ...
$ data : list()
> wdir <- "C:/Users/vlado/docs/papers/2023/HiTeC/London/slides/BACI/95-21"
> setwd(wdir)
> library(jsonlite)
> MN <- fromJSON("WTsect.json")
> str(MN)
> source("https://raw.githubusercontent.com/bavla/ibm3m/master/multiway/MWnets.R")