check_biorxiv.R

## How many new articles have been deposited in biorxiv?  When new
## papers get uploaded to biorxiv, this information is sent to
## crossref.  So, we can use the excellent rcrossref package from the
## ropensci team to get information.
## Note, this does not include revised versions of papers.
##
## Thanks to Scott Chamberlain for providing the magic lines of code to
## grab the information efficiently from crossref.

## Note also there is some code below to grab the infomration from
## http://www.cshsymposium.com/biorxiv/show_all.php
## and on that site you can also generate similar graphs.


require(rcrossref)
system.time(
  res <- cr_prefixes(prefixes = "10.1101", works = TRUE, limit = 1000, cursor = "*", cursor_max = 5000,
                     filter = list(type = "report"), .progress = "text")
)


## colours taken from http://colorbrewer.org
col1 = '#f1eef6'
col2 = '#bdc9e1'
col3 = '#74a9cf'
col4 = '#0570b0'

## note problem with histogram colour changing...
## http://stackoverflow.com/questions/5649600/axis-color-of-date-histogram-in-r
d = as.Date(res$data$created)
##pdf(file="biorxiv_deposits.pdf", width=7, height=4)
svg(file="biorxiv_deposits.svg", width=7, height=4)
par(mar=c(4.5, 4.5, 0, 0.4), cex.axis=0.8)
hist(d, "months", format="%y-%m",main='',
     ylab='First submission', xlab='Date',
     axes=F,
     col=c(rep(col1, 2),
           rep(col2, 12),
           rep(col3, 12),
           rep(col4, 6)                 #increase post June 2016...
           ),
     freq=TRUE, las=2, ylim=c(0,400))
Axis(d,col="black", side=1)
axis(2, col="black", las=1)
dev.off()

q()



### old code below

cr_prefixes(prefixes = c('10.1101'))
cr_cn(dois="10.1126/science.169.3946.635", format="text")
cr_cn(dois="10.1101/045104", format="text")
## graph at: http://www.cshsymposium.com/biorxiv/usage_monthly.php

## screen scrape via: http://www.cshsymposium.com/biorxiv/show_all.php

res <- cr_prefixes(prefixes = "10.1101", works = TRUE, filter = list(type = "report"), offset=1000, limit=1000)
res$data$DOI

cr_cn(dois="10.1101/012799", format="text")


get_chunk = function(offset=0, limit=20) {
  res <- cr_prefixes(prefixes = "10.1101", works = TRUE,
                     filter = list(type = "report"),
                     offset=offset, limit=limit)
  dois = res$data$DOI
  res = cr_works(dois=dois)
  l = data.frame(doi=dois, created=res$data$created)
  l
}


get_chunks = function(offset=0) {
  f = data.frame()
  limit = 1000
  looking = TRUE
  while (looking) {
    f2 = get_chunk(offset, limit)
    offset = offset + limit
    print(offset)
    f = rbind(f, f2)
    if (nrow(f2) < limit) {
      looking = FALSE
    }
  }
  f
}
  
f3 = get_chunks()

dates = as.Date(f3$created)
hist(dates, "months", format="%d %b")
hist(dates, "months", format="%d %b %y", freq=T)
f2 = get_chunk(offset=3000)


######################################################################

## w3m -dump -cols 999  http://www.cshsymposium.com/biorxiv/show_all.php > o2.txt

dat = readLines("o2.txt")

dat = dat[-(1:6)]

## remove blank lines.
g = grep("^$", dat)
dat = dat[-g]
## find last line "Total Articles, Unique: "
g = grep("^Total Articles, Unique", dat)

## throw away tail.
dat = dat[-(g:length(dat))]

dois = substring(dat, 7, 20)
dates = as.Date(substring(dat, 28, 37))
hist(dates, "months", format="%d-%b-%y", freq=TRUE)

frictionless-application

* Application for Frictionless data tool fund

Stephen Eglen, 2020-05-16