2018-10-30

Week 31 - R and Package download stats

Anonymized package and R language downloads from the RStudio CRAN mirror.

• r-downloads.csv - R language downloads from RStudio CRAN mirror on last TidyTuesday for October 23, 2018.
• r_downloads_year.csv - A year's worth of R language downloads from RStudio CRAN mirror between October 20, 2017 and October 20, 2018.

Data Dictionary

Header	Description
date	date of download (y-m-d)
time	time of download (in UTC)
size	size (in bytes)
version	R release version
os	Operating System
country	Two letter ISO country code.
ip_id	Anonymized daily ip code (unique identifier)

Package-level downloads

Package downloads end up with MUCH larger metadata files and are a bit unwieldy to work with, but if you want to play around with them you can use a few ways to easily check!

cran-logs.rstudio.com

• .csv.gz files (can be read directly into R via readr::read_csv()
• Data back to Oct 2012
• Package and R-language downloads (anonymized)

Here's an easy example/way to get all the URLs in R

# Set your range of dates
start <- as.Date('2012-10-01')
today <- as.Date('2018-10-27')

all_days <- seq(start, today, by = 'day')

year <- as.POSIXlt(all_days)$year + 1900

# combine dates into a character vector of dates
urls <- paste0('http://cran-logs.rstudio.com/', year, '/', all_days, '.csv.gz')

# Read directly into memory as dataframe
urls %>%
  map_dfr(read_csv)
  
# Can use download.file to download instead of read into memory

cranlogs

• Extremely easy to use and small data files
• Let's you download by specific packages or look over small time-frames to see what were the most popular packages.
• Con: Aggregated data for the most part

installr::download_RStudio_CRAN_data

• Downloads data from RStudio CRAN anonymized logs via API call
• Has download and read function built in for working with CRAN log data