output: github_document

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

bqschol

The goal of bqschol is to provide an interface to SUB Göttingen's big scholarly datasets stored on Google Big Query.

This package is of internal use.

Installation

You can install the development version from GitHub with:

# install.packages("remotes")
remotes::install_github("njahn82/bqschol")

Initialize the connection

Connect to dataset with Unpaywall snapshots

library(bqschol)
my_con <- bqschol::bgschol_con(
  dataset = "cr_history",
  path = "~/hoad-private-key.json")

You need to have a service account token to make use of this package!

Table functions

The package provides wrapper for the most common table operations

• bgschol_list(): List tables
• bgschol_tbl(): Access tables with
• bgschol_query(): Perform of a SQL query and retrieve results
• bgschol_execute(): Execute a SQL query on the database

Let's start by listing all Crossref snapshots on SUB Göttingen's Big Query project

bgschol_list(my_con)

We can determine the top publisher by type as of April 2018. Note that we only stored Crossref records published later than 2007.

cr_instant_df <- bgschol_tbl(my_con, table = "cr_apr18")
cr_instant_df %>%
    #top publishers
    dplyr::group_by(publisher) %>%
    dplyr::summarise(n = dplyr::n_distinct(doi)) %>%
    dplyr::arrange(desc(n)) 

For more complex tasks, we use SQL.

cc_query <- c("SELECT
  publisher,
  COUNT(DISTINCT(DOI)) AS n
FROM
  `api-project-764811344545.cr_history.cr_apr18`,
  UNNEST(license) AS license
WHERE
  REGEXP_CONTAINS(license.URL, 'creativecommons')
GROUP BY
  publisher
ORDER BY
  n DESC
LIMIT
  10")
bgschol_query(my_con, cc_query)

bgschol_execute() is when new tables shall be created or dropped in Big Query.