xdata_positions.R

library(countrycode)
library(tidyverse)
library(rio)

holder <- list()

### (0) PartyFacts lookup
pf <- read.csv("data/partyfacts-external-parties.csv")

### (1) Manifesto Project
### Accessed through a library
library(manifestoR)
marpor_apikey_locn <- "~/Dropbox/tmp/manifesto_apikey.txt"
mp_setapikey(key.file = marpor_apikey_locn)

marpor <- mp_maindataset()

### There is a very odd error for Norway, 1989, which arises because
### we do not know the length of the manifestos. We'll insert the
### shortest manifesto we know of from Norway.
marpor$total[which(is.na(marpor$total) & marpor$countryname == "Norway" &
                   marpor$edate == "1989-09-11")] <-
    min(marpor$total[which(marpor$countryname == "Norway")], na.rm = T)

marpor$left_right <- mp_scale(marpor, scalingfun = logit_rile)

marpor <- merge(marpor,
                pf %>%
                filter(dataset_key == "manifesto") %>%
                dplyr::select(dataset_key, dataset_party_id, partyfacts_id),
                by.x = "party",
                by.y = "dataset_party_id",
                all.x = TRUE,
                all.y = FALSE)

marpor$iso3c <- countrycode(marpor$countryname,
                            "country.name",
                            "iso3c")

marpor$year <- as.numeric(format(marpor$edate, format = "%Y"))

holder[["manifesto"]] <- marpor %>%
    dplyr::select(left_right = rile,
                  partyfacts_id,
                  iso3c, year) %>%
    mutate(source = "manifesto")


### (2) CHES
ches <- read.csv("data/1999-2019_CHES_dataset_means.csv") %>%
    mutate(left_right = as.numeric(as.character(lrecon)),
           iso3c = NA)

ches <- merge(ches,
              pf %>%
              filter(dataset_key == "ches") %>%
              dplyr::select(dataset_key, dataset_party_id, partyfacts_id),
              by.x = "party_id",
              by.y = "dataset_party_id",
              all.x = TRUE,
              all.y = FALSE)
              
holder[["ches"]] <- ches %>%
    dplyr::select(left_right,
                  partyfacts_id,
                  iso3c, year) %>%
    mutate(source = "ches")


### (3) V-Party
vparty_file <- "data/V-Dem-CPD-Party-V1.rds"

vparty <- readRDS(vparty_file) %>%
    dplyr::select(partyfacts_id = pf_party_id,
                  year,
                  left_right = v2pariglef) %>%
    mutate(source = "vparty")

holder[["vparty"]] <- vparty

### (4) DBPI
## dpi <- rio::import("data/DPI2015.dta") %>%
##     mutate(iso3c = countrycode(countryname,
##                                "country.name",
##                                "iso3c"))


### (5) Huber-Inglehart

### (6) Kitschelt (2013)
kitschelt <- rio::import("data/partylevel_20130907.dta") %>%
    dplyr::select(iso3c = ccodewb, partynum, left_right = dw) %>%
    mutate(dataset_party_id = paste(iso3c, partynum, sep = "-"),
           year = 2010)


kitschelt <- merge(kitschelt,
              pf %>%
              filter(dataset_key == "kitschelt") %>%
              dplyr::select(dataset_key, dataset_party_id, partyfacts_id),
              by = "dataset_party_id",
              all.x = TRUE,
              all.y = FALSE)
              
holder[["kitschelt"]] <- kitschelt %>%
    dplyr::select(left_right,
                  partyfacts_id,
                  iso3c, year) %>%
    mutate(source = "kitschelt")

### (7) PPMD
### ppmd <- rio::import("data/ppmd_summary_data.por")

### (8) Comparative Candidates?

### (9) Coppedge
###

### (10) Global Party Survey
gps <- read.csv("data/Global Party Survey by Party SPSS V2_1_Apr_2020-2.csv") %>%
    dplyr::select(ID_GPS,
                  iso3c = ISO,
                  left_right = V4_Scale,
                  year = Elec_year)

gps <- merge(gps,
              pf %>% 
              filter(dataset_key == "gps") %>%
              dplyr::select(dataset_key, dataset_party_id, partyfacts_id),
              by.x = "ID_GPS",
              by.y = "dataset_party_id",
             all.x = TRUE,
             all.y = FALSE) 

holder[["gps"]] <- gps %>%
    dplyr::select(-ID_GPS) %>%
    mutate(source = "gps")

### (11) Janda

### (12) KUREP
## kurep <- rio::import("data/ceeandwe2007elitesurveyspublic06-24-2013.dta") %>%
##     mutate(countryname = characterize(country))

### (13) Laver/Hunt

### (14) Morgan (1976)

### (15) POPPA

### (16) PPLA

### (17) Ray

### (18) Thomas


dat <- bind_rows(holder) %>%
    pivot_wider(id_cols = c(year, partyfacts_id),
                values_from = left_right,
                names_from = source,
                values_fn = mean)

panel.cor <- function(x, y, digits = 2, prefix = "", cex.cor, ...)
{
    usr <- par("usr"); on.exit(par(usr))
    par(usr = c(0, 1, 0, 1))
    r <- abs(cor(x, y, use = "pairwise"))
    txt <- format(c(r, 0.123456789), digits = digits)[1]
    txt <- paste0(prefix, txt)
    if(missing(cex.cor)) cex.cor <- 0.8/strwidth(txt)
    text(0.5, 0.5, txt, cex = cex.cor * r)
}

png(file = "cross_data_cors.png")
pairs(dat[,c("manifesto", "ches", "vparty", "kitschelt", "gps")],
      labels = c("Manifesto Project\n(logit scaling)", "CHES", "V-Party",
                 "Kitschelt (2013)", "Global Party Survey"),
      lower.panel = panel.smooth, upper.panel = panel.cor)


dev.off()