rcheology

A data package which lists every command in base R packages since R version 1.0.1.

The latest R version covered is 4.0.4.

You can view the data online in a Shiny app.

Installing

From CRAN:

install.packages('rcheology')

Where the data comes from

Versions 4.0.0 and up are installed from the CRAN apt repositories for Ubuntu Bionic.

Versions 3.0.1 to 3.6.3 are installed from the CRAN apt repositories for Ubuntu Trusty Tahr. Version 3.5.0 and up use a special repository.

Versions 2.5.1 to 3.0.0 are built from source on Ubuntu Lucid Lynx.

Versions 1.2.3 to 2.4.1 are mostly built from source on Debian Sarge.

Versions 1.0.1 to 1.2.2 (and a couple of later versions) are built from source on Debian Woody.

Results are found from running ls on all installed packages from a minimal installation. Recommended packages are not included.

The Rversions data frame lists versions of R and release dates.

Do it yourself

• Install docker.
• ./control build builds the images. Or get them from https://hub.docker.com/r/dash2/rcheology/.
• ./control run runs the images to build/install R and extract data
• ./control gather gets CSV files from the containers
• ./control write puts CSV files into a data frame and stores it in the package

The data

library(rcheology)
data("rcheology")

rcheology[rcheology$name == "kmeans" & rcheology$Rversion %in% c("1.0.1", "1.9.0", "2.1.0", "3.0.2", "3.2.0", "4.0.2"), ]
#>        package   name Rversion    type exported    class generic
#> 191405     mva kmeans    1.0.1 closure     TRUE     <NA>   FALSE
#> 209743   stats kmeans    1.9.0 closure     TRUE function   FALSE
#> 209747   stats kmeans    2.1.0 closure     TRUE function   FALSE
#> 209786   stats kmeans    3.0.2 closure     TRUE function   FALSE
#> 209791   stats kmeans    3.2.0 closure     TRUE function   FALSE
#> 209816   stats kmeans    4.0.2 closure     TRUE function   FALSE
#>                                                                                                                              args
#> 191405                                                                                                (x, centers, iter.max = 10)
#> 209743                                                                                                (x, centers, iter.max = 10)
#> 209747                  (x, centers, iter.max = 10, nstart = 1, algorithm = c("Hartigan-Wong",     "Lloyd", "Forgy", "MacQueen"))
#> 209786   (x, centers, iter.max = 10, nstart = 1, algorithm = c("Hartigan-Wong",     "Lloyd", "Forgy", "MacQueen"), trace = FALSE)
#> 209791 (x, centers, iter.max = 10L, nstart = 1L, algorithm = c("Hartigan-Wong",     "Lloyd", "Forgy", "MacQueen"), trace = FALSE)
#> 209816 (x, centers, iter.max = 10L, nstart = 1L, algorithm = c("Hartigan-Wong",     "Lloyd", "Forgy", "MacQueen"), trace = FALSE)

Latest changes:

suppressPackageStartupMessages(library(dplyr))

r_penultimate <- sort(package_version(unique(rcheology::rcheology$Rversion)), 
      decreasing = TRUE)
r_penultimate <- r_penultimate[2]

r_latest_obj <- rcheology %>% filter(Rversion == r_latest)
r_penult_obj <- rcheology %>% filter(Rversion == r_penultimate)

r_introduced <- anti_join(r_latest_obj, r_penult_obj, by = c("package", "name"))

r_introduced
#> [1] package  name     Rversion type     exported class    generic  args    
#> <0 rows> (or 0-length row.names)

Base functions over time:

library(ggplot2)

rvs <- rcheology$Rversion     %>% 
      unique()                %>% 
      as.package_version()    %>% 
      sort() %>% 
      as.character()

major_rvs <- grep(".0$", rvs, value = TRUE)
major_rv_dates <- Rversions$date[Rversions$Rversion %in% major_rvs]
major_rvs <- gsub("\\.0$", "", major_rvs)

rch_dates <- rcheology %>% left_join(Rversions, by = "Rversion")
ggplot(rch_dates, aes(date, group = package, fill = package), colour = NA) + 
      stat_count(geom = "area") + 
      theme(axis.text.x = element_text(angle = 45, hjust = 1, size = 8)) + 
      # ggthemes::scale_fill_gdocs() +
      scale_x_date(breaks  = major_rv_dates, labels = major_rvs) + 
      xlab("Version") + ylab("Function count") + 
      theme(legend.position = "top")

An alternative view:

ggplot(rch_dates, aes(date, fill = "orange")) + 
      stat_count(geom = "area") + 
      scale_x_date(breaks  = major_rv_dates, labels = major_rvs) + 
      theme(axis.text.x = element_text(angle = 45, hjust = 1, size = 8)) + 
      xlab("Version") + ylab("Function count") + 
      facet_wrap(~package, scales = "free_y", ncol = 2) +
      theme(legend.position = "none")