changes_metrics_small_multiples.R

library(dplyr)
library(tidyr)
library(ggplot2)
library(gridExtra)
library(grid)

data20162017 <- read.csv("toshiny_altm_2016_2017.csv", stringsAsFactors = FALSE)

data20162017 <- data20162017 %>% 
  mutate(gap_cites = Citations2017 - Citations,
         gap_tweets = Twitter2017 - Twitter,
         gap_mend = Mendeley2017 - Mendeley)

data20162017_gathered <- gather(data20162017, metrics, value, Citations, Citations2017, Mendeley, Mendeley2017, Twitter, Twitter2017)

data20162017_gathered <- data20162017_gathered %>%
  mutate(Date = ifelse(grepl('2017', metrics),
                       format(as.Date("2017-05-30", format="%Y-%m-%d"),"%Y-%m-%d"),
                       format(as.Date("2016-05-13", format="%Y-%m-%d"),"%Y-%m-%d")))

data20162017_gathered$metrics <- gsub("([^0-9]*)[0-9]+","\\1",data20162017_gathered$metrics)
data20162017_gathered$Date <- as.Date(data20162017_gathered$Date, origin="1970-01-01")
data20162017_gathered$value <- as.numeric(data20162017_gathered$value)
data20162017_gathered$DOI <- gsub("/", "/\n", data20162017_gathered$DOI)

data20162017_gathered$value[is.na(data20162017_gathered$value)] <- 0

# https://blog.dominodatalab.com/visualizing-homeownership-in-the-us-using-small-multiples-and-r/
opts <- theme(
  panel.background = element_rect(fill="white"),
  panel.border = element_rect(colour="black", fill=NA),
  axis.line = element_blank(),
  axis.ticks = element_blank(),
  panel.grid.major = element_blank(),
  panel.grid.minor = element_blank(),
  axis.text = element_text(colour="gray25", size=15, angle=90, hjust=1),
  axis.title = element_blank(),
  text = element_text(size=12),
  legend.key = element_blank(),
  legend.position = "none",
  legend.background = element_blank(),
  plot.title = element_text(size = 45))

vplayout <- function(x, y) viewport(layout.pos.row = x, layout.pos.col = y)

doschool <- function(data = NULL, sc = NULL, m = NULL) {
  
  m_name <- if(m=="gap_cites") "citations" else if(m=="gap_tweets") "tweets" else "Mendeley readers"
  
  message("Now printing ", sc, " sorted by diff in ", m_name)
  flush.console()
  
  s <- data %>% 
    filter(School == sc) %>% 
    arrange_(.dots = paste0("desc(", m, ")"))
  
  dois <- unique(s$DOI)
  
  if (length(dois)<=50) {
    dois <- dois
  } else {
    dois <- dois[1:100]
  }
  
  grid.newpage()
  
  if ( s[1, "School"]=="ARTS" ) {
    png(file=paste0("altm_", m, "_", sc, ".png"), height = 1500, width = 3000)
    pushViewport(viewport(layout = grid.layout(7, 10, heights=c(.3,.1,rep(1,5)), widths=c(1.1, rep(1,9)))))
  } else {
    png(file=paste0("altm_", m, "_", sc, ".png"), height = 3500, width = 4000)
    pushViewport(viewport(layout = grid.layout(12, 10, heights=c(.3,.1,rep(1,10)), widths=c(1.1, rep(1,9)))))
  }
  
  grid.text("WoS citations, Mendeley readers, and tweets 2016-2017", 
            gp=gpar(fontsize=50, col="gray40"), 
            vp = viewport(layout.pos.row = 1, 
                          layout.pos.col = 1:10))
  grid.text(paste0("Aalto University, School of ", sc, ". Max 100 items published 2013-2015, sorted by changes in ",  m_name, " between 2016-05-13 and 2017-06-01."), 
            gp=gpar(fontsize=30, col="gray40"), 
            vp = viewport(layout.pos.row = 2, 
                          layout.pos.col = 1:10))
  
  for (i in 1:length(dois))
  {
    datap <- subset(s, DOI==dois[i])
    p <- ggplot(datap, aes(Date, value))+
      geom_rect(data = datap, 
                aes(fill = factor(metrics)), 
                xmin = -Inf, xmax = Inf, ymin = -Inf, ymax = Inf, alpha = NA)+
      scale_fill_brewer(palette="Set2")+
      geom_line(size=1.5)+
      facet_wrap(DOI ~ metrics)+
      {if ( datap[1,"School"]=="ARTS" )  { scale_y_continuous(expand = c(0, 0), limits = c(0,190)) } else  { scale_y_log10(limits = c(1,1e3)) }}+
      scale_x_date(date_breaks = "1 year", date_labels = "%Y")+
      opts 
    if (i%%10!=1) p <- p+theme(axis.text.y = element_blank())
    if (datap[1,"School"]=="ARTS") {
      if (i<=40)    p <- p+theme(axis.text.x = element_blank())
    } else {
      if (i<=90)    p <- p+theme(axis.text.x = element_blank())
    }
    print(p, vp = vplayout(floor((i-1)/10)+3, i%%10+(i%%10==0)*10))
  }
  dev.off() 
}

metrics <- c("gap_cites", "gap_tweets", "gap_mend")
schools <- sort(unique(data_gathered$School))

for (i in 1:length(schools)) {
  for (j in 1:length(metrics)) {
    doschool(data20162017_gathered, schools[i], metrics[j])
  }
}