osm_review_figures.Rmd

---
title: "OSM community literature review"
always_allow_html: yes
output:
  html_document:
    df_print: paged
    fig_caption: yes
  pdf_document: 
    fig_caption: yes
---

### Setting up the environment

```{r functions, include=FALSE}
# A function for captioning and referencing images
fig <- local({
    i <- 0
    ref <- list()
    list(
        cap=function(refName, text) {
            i <<- i + 1
            ref[[refName]] <<- i
            paste("Figure ", i, ": ", text, sep="")
        },
        ref=function(refName) {
            ref[[refName]]
        })
})
```


```{r setup,include=F}
if (!require('ggplot2')) {
  install.packages('ggplot2')
  library('ggplot2')
}
if (!require('circlize')) {
  install.packages('circlize')
  library('circlize')
}
if (!require('grid')) {
  install.packages('grid')
  library('grid')
}
if (!require('gridExtra')) {
  install.packages('gridExtra')
  library('gridExtra')
}
if (!require('ggmap')) {
  install.packages('ggmap')
  library('ggmap')  
}
if (!require('maps')) {
  install.packages('maps')
  library('maps')  
}
if (!require('mapdata')) {
  install.packages('mapdata')
  library('mapdata')  
}
if (!require('sf')) {
  install.packages('sf')
  library('sf')  
}
if (!require('rgeos')) {
  install.packages('rgeos')
  library('rgeos')  
}
if (!require('rnaturalearth')) {
  install.packages('rnaturalearth')
  library('rnaturalearth')  
}
if (!require('rnaturalearthdata')) {
  install.packages('rnaturalearthdata')
  library('rnaturalearthdata')  
}
if (!require('plyr')) {
  install.packages('plyr')
 library('plyr')  
}
if (!require('dplyr')) {
  install.packages('dplyr')
 library('dplyr')  
}
if (!require('reshape2')) {
  install.packages('reshape2')
  library('reshape2')
}
if (!require('data.table')) {
  install.packages('data.table')
  library('data.table')
}
if (!require('tm')) {
  install.packages('tm')
  library("tm")
}
if (!require('SnowballC')) {
  install.packages('SnowballC')
  library("SnowballC")
}
if (!require('wordcloud2')) {
  install.packages('wordcloud2')
  library("wordcloud2")
}
if (!require('RColorBrewer')) {
  install.packages('RColorBrewer')
  library("RColorBrewer")
}
if (!require('waffle')) {
  install.packages('waffle')
  library("waffle")
}

setwd(dirname(rstudioapi::getActiveDocumentContext()$path))

```

Output not included. Check markdown.

### Read an transform data

```{r}
raw_papers <- read.csv('OSM_Paper_Review.csv', header=T, stringsAsFactors=F)

```

Convert data frame to list, also split comma and semicolon separated values to vectors.

```{r}
paper_list <- split(raw_papers, seq(nrow(raw_papers)))

for (paper in seq_along(paper_list)) {
  # Use an ugly, ineffective regex to split string by 2 delimiters.
  paper_list[[paper]]$Authors..Geography..countries. <- strsplit(paper_list[[paper]]$Authors..Geography..countries., '\\,|\\;|\\, | \\; ')
  paper_list[[paper]]$Authors..Geography..continents. <- strsplit(paper_list[[paper]]$Authors..Geography..continents., '\\,|\\;|\\, | \\; ')
  paper_list[[paper]]$Study.Area.Geography..countries. <- strsplit(paper_list[[paper]]$Study.Area.Geography..countries., '\\,|\\;|\\, | \\; ')
  paper_list[[paper]]$Evidence.of.engagement <- strsplit(paper_list[[paper]]$Evidence.of.engagement, '\\,|\\;|\\, | \\; ')
  paper_list[[paper]]$Perspective.on.the.community <- strsplit(paper_list[[paper]]$Perspective.on.the.community, '\\,|\\;|\\, | \\; ')
}

```

### Report

Collect countries, continents and categories into vectors

```{r}
country_author <- c()
continent_author <- c()
country_study <- c()
engagement <- c()
perspective <- c()

# For all paper insert combine vectors. Also trim leading whitespaces introduced by the lazy regex
for (paper in seq_along(paper_list)) {
#  print(paste(paper, trimws(paper_list[[paper]]$study_area_country_fix[[1]])))
 country_author <- c(country_author, trimws(paper_list[[paper]]$Authors..Geography..countries[[1]]))
 continent_author <- c(continent_author, trimws(paper_list[[paper]]$Authors..Geography..continents.[[1]]))
 country_study <- c(country_study, trimws(paper_list[[paper]]$Study.Area.Geography..countries.[[1]]))
 engagement <- c(engagement, trimws(paper_list[[paper]]$Evidence.of.engagement[[1]]))
 perspective <- c(perspective, trimws(paper_list[[paper]]$Perspective.on.the.community[[1]]))
}

```

Some countries do not match the Natural Earth data. Correct them manually for now

```{r}
# Load Natural Earth data
world <- ne_countries(scale = "medium", returnclass = "sf")

# Check which countries can't be matched to Natural Earth (TODO: fix them in the spreadsheet)
unique(country_author)[unique(country_author) %in% world$admin == F]
unique(country_study)[unique(country_study) %in% world$admin == F]


# Fix them manually
country_author <- replace(country_author, country_author == 'Serbia', 'Republic of Serbia')
country_author <- replace(country_author, country_author == 'Hong Kong', 'Hong Kong S.A.R.')
country_author <- replace(country_author, country_author == 'United States', 'United States of America')

country_study <- replace(country_study, country_study == 'Serbia', 'Republic of Serbia')
country_study <- replace(country_study, country_study == 'United States', 'United States of America')
country_study <- replace(country_study, country_study == 'Hong Kong', 'Hong Kong S.A.R.')
country_study <- replace(country_study, country_study == 'Tanzania', 'United Republic of Tanzania')
country_study <- replace(country_study, country_study == 'Madgascar', 'Madagascar')

for (paper in seq_along(paper_list)) {
  paper_list[[paper]]$Authors..Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Authors..Geography..countries.[[1]], paper_list[[paper]]$Authors..Geography..countries.[[1]] == 'Serbia', 'Republic of Serbia')
  paper_list[[paper]]$Authors..Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Authors..Geography..countries.[[1]], paper_list[[paper]]$Authors..Geography..countries.[[1]] == 'Hong Kong', 'Hong Kong S.A.R.')
  paper_list[[paper]]$Authors..Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Authors..Geography..countries.[[1]], paper_list[[paper]]$Authors..Geography..countries.[[1]] == 'United States', 'United States of America')
  
  paper_list[[paper]]$Study.Area.Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Study.Area.Geography..countries.[[1]], paper_list[[paper]]$Study.Area.Geography..countries.[[1]] == 'United States', 'United States of America')
  paper_list[[paper]]$Study.Area.Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Study.Area.Geography..countries.[[1]], paper_list[[paper]]$Study.Area.Geography..countries.[[1]] == 'Serbia', 'Republic of Serbia')
  paper_list[[paper]]$Study.Area.Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Study.Area.Geography..countries.[[1]], paper_list[[paper]]$Study.Area.Geography..countries.[[1]] == 'Tanzania', 'United Republic of Tanzania')
  paper_list[[paper]]$Study.Area.Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Study.Area.Geography..countries.[[1]], paper_list[[paper]]$Study.Area.Geography..countries.[[1]] == 'Hong Kong', 'Hong Kong S.A.R.')
  paper_list[[paper]]$Study.Area.Geography..countries.[[1]] <-  replace(paper_list[[paper]]$Study.Area.Geography..countries.[[1]], paper_list[[paper]]$Study.Area.Geography..countries.[[1]] == 'Madgascar', 'Madagascar')
}

```

#### Create maps

Assign frequencies to countries

```{r}
# Assign frequencies to countries
world$country_author <- as.numeric(sapply(world$admin, FUN=function(country) { return(country_author_table[country_author_table$country_author==country,]$Freq)}))
world$country_study <- as.numeric(sapply(world$admin, FUN=function(country) { return(study_area_table[study_area_table$country_study==country,]$Freq)}))

world$country_author[is.na(world$country_author)] <- 0
world$country_study[is.na(world$country_study)] <- 0

# Extract centroids for labels
world_points <- st_centroid(world)
world_points <- cbind(world, st_coordinates(st_centroid(world$geometry)))
```

Plot country of authors

```{r, fig.cap="Home countries of authors"}
png('home_country_authors.png', height=1200, width=2400, unit='px')
ggplot(data=world) + geom_sf(aes(fill=country_author)) + theme_bw() + coord_sf(crs = "+proj=robin +lon_0=0 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs ") + scale_fill_gradient(low='white', high='red') + ggtitle('Home country of authors') + labs(fill='') + theme(title=element_text(size=36), legend.text=element_text(size=28), legend.key.height=unit(3, 'cm'), panel.grid=element_line(size=2)) 
dev.off()
```

Plot study areas (excluding "global")

```{r, fig.cap="Countries receiving the most research attention"}
png('study_area_country.png', height=1200, width=2400, unit='px')

ggplot(data=world) + geom_sf(aes(fill=country_study)) + theme_bw() + coord_sf(crs = "+proj=robin +lon_0=0 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs ") + scale_fill_gradient(low='white', high='red') + ggtitle('Focus of study') + labs(fill='') + theme(title=element_text(size=36), legend.text=element_text(size=28), legend.key.height=unit(3, 'cm')) 
dev.off()
```

Assign focus continent for chord diagram

```{r}
for (paper in seq_along(paper_list)) {
  continents <- c()
  for (country in paper_list[[paper]]$Study.Area.Geography..countries.[[1]]) {

    continents <- unique(c(continents, world[world$admin==country,]$continent))
    
    if (!is.na(country) & country == 'Global') {
      continents <- unique(c(continents, 'Global'))
     }
    if (!is.na(country) & country == 'Europe') {
      continents <- unique(c(continents, 'Europe'))
    }
    if (!is.na(country) & country == 'Asia') {
      continents <- unique(c(continents, 'Asia'))
    }
    #if (is.na(country)) {
    #  continents <- unique(c(continents, 'not applicable'))
    #}
  }

  paper_list[[paper]]$Study.Area.Geography..continents. <- list(continents)
}

study_continent_list <- c()
global_ids <- c()

fromto <- data.frame(from_continent=character(), to_continent=character(), value=numeric())
i <- 0
sum_values <- 0
for (paper in seq_along(paper_list)) {
  paper_value <- 0
  #i <- i + 1
  for (from_continent in paper_list[[paper]]$Authors..Geography..continents.[[1]]) {
    if (from_continent=='Australia & Oceania') {
      from_continent <- 'Oceania'
    }
    if (from_continent=='Australia') {
      from_continent <- 'Oceania'
    }
    
    for (study_continent in paper_list[[paper]]$Study.Area.Geography..continents.[[1]]) {
      study_continent_list <- c(study_continent_list, study_continent)
      if (!is.na(study_continent) & study_continent=='Global') {
        #print(paste(paper_list[[paper]]$id, 'Global'))
        global_ids <- c(global_ids, paper_list[[paper]]$id)
        }
      if (nrow(fromto[fromto$from_continent==from_continent & fromto$study==study_continent,]) == 0) {
        if (is.na(study_continent)) next
        fromto <- rbind(fromto, data.frame(from_continent, study_continent, value = 1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])))
        ## Check value of one paper
        paper_value <- paper_value + 1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])
        sum_values <- sum_values +   1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])
      } else {
        if (is.na(study_continent)) next
        fromto[fromto$from_continent==from_continent & fromto$study==study_continent,]$value <- fromto[fromto$from_continent==from_continent & fromto$study==study_continent,]$value + 1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])
        paper_value <- paper_value + 1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])
        sum_values <- sum_values +   1/length(paper_list[[paper]]$Authors..Geography..continents.[[1]]) / length(paper_list[[paper]]$Study.Area.Geography..continents.[[1]])
      }
    }
  }
  i <- i + paper_value
}

fromto$from_continent <- as.character(fromto$from_continent)
fromto$from_continent <-  replace(fromto$from_continent, fromto$from_continent == 'Australia & Oceania', 'Oceania')
fromto$from_continent <-  replace(fromto$from_continent, fromto$from_continent == 'Australia', 'Oceania')


fromto <- fromto[!(fromto$from_continent=='not applicable' | fromto$study_continent=='not applicable'), ]

fromto
```

```{r}
svg('from_to_studies_directional.svg')#, pointsize=12)

chordDiagram(fromto, grid.col=brewer.pal(7, 'Set1'), annotationTrack = "grid", preAllocateTracks = 1, directional=1, direction.type=c('arrows','diffHeight'), diffHeight = 0.05, link.sort = TRUE, link.largest.ontop = TRUE, target.prop.height=2)
circos.trackPlotRegion(track.index = 1, panel.fun = function(x, y) {
  xlim = get.cell.meta.data("xlim")
  ylim = get.cell.meta.data("ylim")
  sector.name = get.cell.meta.data("sector.index")
  circos.text(mean(xlim), ylim[1] + 2, sector.name, cex=2.5, facing = "clockwise", niceFacing = TRUE, adj = c(0, 0.5))
  circos.axis(h = "top", labels.cex = 0.5, major.tick.length = 3, sector.index = sector.name, track.index = 2)
}, bg.border = NA)
dev.off()
```