A friend of mine, Dr. Abdullah Aydogan (https://twitter.com/abdaydgn), has asked me this morning if it is possible to pull the data for the 2019 local elections in Turkey. The only accessible information is through the Anadolu Agency (https://www.aa.com.tr/en) because the official election organization’s website (http://www.ysk.gov.tr/) has not been working for a while. Here, I provide the code I used to scrap data from the Anadolu Agency, only available source for election results.
A friend of mine, @abdaydgn, has asked me this morning if it is possible to pull the data for the 2019 local elections in Turkey. The only accessible information is through the Anadolu Agency because the official election organization’s website (YSK) has not been working for a while.
After doing some detective work about where the data is stored, I fount the the link http://secim.aa.com.tr/Handlers/IlBelediyeBaskanligi.ashx for the city results.
For the districts within each city, the link is http://secim.aa.com.tr/Handlers/IlceBelediye.ashx?il=citycode, where you have to replace a number in place of citycode for each city. For instance, city code for Istanbul is 34. So, the link http://secim.aa.com.tr/Handlers/IlceBelediye.ashx?il=34 will return the results for all districts of Istanbul.
Once the data source is found, the rest is a little bit JSON data processing. I don’t have time now to explain the code, but I will post the data files and the code below. Maybe, I can come back and update this post later to put more explanation about the code.
Long Format
Wide Format
# Install the following packages
install.packages("jsonlite")
install.packages("mefa")
install.packages("car")
install.packages("plyr")
require(jsonlite)
require(mefa)
require(car)
require(plyr)
##########################################################################
#
# Pulling Data for District Results
#
##########################################################################
city.code <- read.csv("il kod.csv")
# This is a 2-column data for each city and corresponding code
# can be downloaded from the following link.
# https://miami.box.com/s/gkwihhqgule7moxoabohwfgh6u74yhkl
secim2019 <- vector("list",81)
for(city in 1:81){
a <- jsonlite::fromJSON(paste0("http://secim.aa.com.tr/Handlers/IlceBelediye.ashx?il=",city))
info <- cbind(a$Ilce$Ad,a$Ilce$Sandik)
sonuc <- vector("list",nrow(info))
for(i in 1:nrow(info)){
n = nrow(a$Ilce$PartiSonuclari$Parti[[i]])
sonuc[[i]] <- cbind(rep(info[i,],n),a$Ilce$PartiSonuclari$Parti[[i]])
}
il <- sonuc[[1]]
for(i in 2:nrow(info)){
il <- rbind(il,sonuc[[i]])
}
il$Il <- as.character(city.code[city,1])
secim2019[[city]] <- il
}
secim2019.all <- secim2019[[1]]
for(i in 2:81){
secim2019.all <- rbind(secim2019.all,secim2019[[i]])
}
secim2019.all$parti.kod <- recode(secim2019.all$Kod,
recodes="1 ='SP';
2 ='BTP';
3 ='TKP';
4 ='VP';
5 ='BBP';
6 ='HudaPar';
7 ='CHP';
8 ='AKP';
9 ='DP';
10 ='MHP';
11 ='IYI';
12 ='HDP';
13 ='DSP';
else ='Bagimsiz'")
secim2019.all$ToplamSandik <- as.numeric(as.character(secim2019.all$ToplamSandik))
secim2019.all$ToplamSecmen <- as.numeric(as.character(secim2019.all$ToplamSecmen))
secim2019.all$AcilanSandikSecmenSayisi <- as.numeric(as.character(secim2019.all$AcilanSandikSecmenSayisi))
secim2019.all$AcilanSandik <- as.numeric(as.character(secim2019.all$AcilanSandik))
secim2019.all$KullanilanOy <- as.numeric(as.character(secim2019.all$KullanilanOy))
secim2019.all$GecerliOy <- as.numeric(as.character(secim2019.all$GecerliOy))
secim2019.all$Kod <- as.numeric(as.character(secim2019.all$Kod))
secim2019.all$Oy <- as.numeric(as.character(secim2019.all$Oy))
secim2019.all$Turnout <- secim2019.all$KullanilanOy/secim2019.all$ToplamSecmen
secim2019.all$GecerliOy.Orani <- secim2019.all$GecerliOy/secim2019.all$KullanilanOy
secim2019.all$Parti.Oran <- secim2019.all$Oy/secim2019.all$GecerliOy
secim2019.all <- secim2019.all[,c("Il","a.Ilce.Ad","ToplamSandik","ToplamSecmen","KullanilanOy",
"GecerliOy","parti.kod","Oy","Turnout","GecerliOy.Orani",
"Parti.Oran")]
colnames(secim2019.all)[2] <- "Ilce"
colnames(secim2019.all)[7] <- "Parti"
write.table(secim2019.all,"Secim2019IlceBelediyeSonuc.csv",row.names=FALSE,sep=",")
# Wide Format
# Aggregate the vote for "Bagimsiz" category for districts with more than
# "Bagimsiz" candidates
secim2019.all.wide <- ddply(secim2019.all,
.(Il, Ilce,ToplamSandik,ToplamSecmen,KullanilanOy,
GecerliOy,Parti,Turnout,GecerliOy.Orani),
summarise,Oy = sum(Oy))
secim2019.all.wide$Parti.Oran <- secim2019.all.wide$Oy/secim2019.all.wide$GecerliOy
secim2019.all.wide <- reshape(secim2019.all.wide,
timevar =c("Parti"),
idvar =c("Il","Ilce","ToplamSandik",
"ToplamSecmen","KullanilanOy",
"GecerliOy","Turnout","GecerliOy.Orani"),
direction="wide")
secim2019.all.wide[is.na(secim2019.all.wide)] <- 0
write.table(secim2019.all.wide,"Secim2019IlceBelediyeSonuc_wide.csv",row.names=FALSE,sep=",")
##########################################################################
#
# Pulling Data for City Results
#
##########################################################################
a <- jsonlite::fromJSON("http://secim.aa.com.tr/Handlers/IlBelediyeBaskanligi.ashx")
info <- cbind(a[[1]][[3]]$Il$Ad,a[[1]][[3]]$Il$Sandik)
sonuc <- vector("list",nrow(info))
for(i in 1:nrow(info)){
n = nrow(a[[1]][[3]]$Il$PartiSonuclari$Parti[[i]])
sonuc[[i]] <- cbind(rep(info[i,],n),a[[1]][[3]]$Il$PartiSonuclari$Parti[[i]])
}
il <- sonuc[[1]]
for(i in 2:nrow(info)){
il <- rbind(il,sonuc[[i]])
}
il$parti.kod <- recode(il$Kod,
recodes="1 ='SP';
2 ='BTP';
3 ='TKP';
4 ='VP';
5 ='BBP';
6 ='HudaPar';
7 ='CHP';
8 ='AKP';
9 ='DP';
10 ='MHP';
11 ='IYI';
12 ='HDP';
13 ='DSP';
else ='Bagimsiz'")
il$ToplamSandik <- as.numeric(as.character(il$ToplamSandik))
il$ToplamSecmen <- as.numeric(as.character(il$ToplamSecmen))
il$AcilanSandikSecmenSayisi <- as.numeric(as.character(il$AcilanSandikSecmenSayisi))
il$AcilanSandik <- as.numeric(as.character(il$AcilanSandik))
il$KullanilanOy <- as.numeric(as.character(il$KullanilanOy))
il$GecerliOy <- as.numeric(as.character(il$GecerliOy))
il$Kod <- as.numeric(as.character(il$Kod))
il$Oy <- as.numeric(as.character(il$Oy))
il$Turnout <- il$KullanilanOy/il$ToplamSecmen
il$GecerliOy.Orani <- il$GecerliOy/il$KullanilanOy
il$Parti.Oran <- il$Oy/il$GecerliOy
colnames(il)[1] <- "IL"
colnames(il)[11] <- "Parti"
il <- il[,c("IL","ToplamSandik","ToplamSecmen","KullanilanOy","GecerliOy","Parti",
"Oy","Turnout","GecerliOy.Orani","Parti.Oran")]
write.table(il,"Secim2019IlBelediyeSonuc.csv",row.names=FALSE,sep=",")
# Wide Format
# Aggregate the vote for "Bagimsiz" category for districts with more than
# "Bagimsiz" candidates
il.wide <- ddply(il,
.(IL,ToplamSandik,ToplamSecmen,KullanilanOy,
GecerliOy,Parti,Turnout,GecerliOy.Orani),
summarise,Oy = sum(Oy))
il.wide$Parti.Oran <- il.wide$Oy/il.wide$GecerliOy
il.wide <- reshape(il.wide,
timevar =c("Parti"),
idvar =c("IL","ToplamSandik",
"ToplamSecmen","KullanilanOy",
"GecerliOy","Turnout","GecerliOy.Orani"),
direction="wide")
il.wide[is.na(il.wide)] <- 0
write.table(il.wide,"Secim2019IlBelediyeSonuc_wide.csv",row.names=FALSE,sep=",")For attribution, please cite this work as
Zopluoglu (2019, April 5). Cengiz Zopluoglu: Scraping Data for 2019 Local Elections in Turkey. Retrieved from https://github.com/czopluoglu/website/tree/master/docs/posts/local-elections-2019-turkey/
BibTeX citation
@misc{zopluoglu2019scraping,
author = {Zopluoglu, Cengiz},
title = {Cengiz Zopluoglu: Scraping Data for 2019 Local Elections in Turkey},
url = {https://github.com/czopluoglu/website/tree/master/docs/posts/local-elections-2019-turkey/},
year = {2019}
}