psych sci author network

psych_sci_author_net.R

library(tidyverse)
library(viridis)
library(knitr)

d <- read_csv("AuthorList.csv")

authors <- d %>%
  group_by(author) %>%
  count

d %>% 
  group_by(year) %>%
  summarise(n_papers = length(unique(article_id)), 
            n_authors = length(author),
            n_authors_per_paper = n_authors / n_papers,
            n_unique_authors = length(unique(author))) %>%
  kable(digits = 2)


unique_authors <- d %>% 
  select(-article_id) %>%
  distinct
  
  

overlaps <- d %>%
  split(.$year) %>%
  map_df(function(x) {
    this_year_authors = unique(x$author)
    overlap <- unique_authors %>% 
      group_by(year) %>% 
      summarise(overlap = sum(author %in% this_year_authors), 
                dice = (2 * overlap) / (length(author) + length(this_year_authors))) %>%
      mutate(source_year = x$year[1])
  }) %>%
  mutate(overlap = ifelse(year == source_year, NA, overlap),
         dice = ifelse(year == source_year, NA, dice))

ggplot(filter(overlaps, year > 2008 & year < 2018, 
              source_year > 2008 & source_year < 2018), 
       aes(x = year, y = source_year, fill = dice)) + 
  geom_tile() + 
  xlim(2009,2018) + 
  scale_fill_viridis() + 
  theme_bw()