How to find out how much of R Core is R

Cleaning out my computer as I get ready to switch to a new one has me running into old gems. So, when I say “Today I learned,” I really mean “I learned this back in December 2021.” 😅

Back then, I gave a talk at Why R? called Packages for Using R With Python, Tableau, and Other Tools. One part of the talk was about how R itself isn’t just made up of R.

I adapted this classic blog post by wrathematics to explore the composition of the R 4.1.2 source package. The post features a script that scans the .R, .c, and .f files in the source, then records the language (R, C, or Fortran) and the number of lines of code in each language to a CSV file. Keep in mind, I have almost no knowledge of Shell (and this was pre-ChatGPT days!), so it took me a bit to adapt the original script from 2011.

shell.sh

outdir="./"

rdir="./R-4.1.2" #eg, ~/R-2.13.1/
cd $rdir/src

for rfile in `find . -type f -name *.R`
do
loc=`wc -l $rfile | sed -e 's/ ./,/' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\///g' -e 's/\///'`
echo "R,$loc"  >> $outdir/r_source_loc.csv
done

for cfile in `find . -type f -name *.c`
do
loc=`wc -l $cfile | sed -e 's/ ./,/' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\///g' -e 's/\///'`
echo "C,$loc"  >> $outdir/r_source_loc.csv
done

for ffile in `find . -type f -name *.f`
do
loc=`wc -l $ffile | sed -e 's/ ./,/' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\//\//g' -e 's/\/[^/]*\///g' -e 's/\///'`
echo "Fortran,$loc"  >> $outdir/r_source_loc.csv
done

The script creates a file called r_source_loc.csv. It shows the number of lines by programming language by script in R 4.1.2. We can read it into R:

library(dplyr)
library(stringr)

r_loc <-
  readr::read_table(here::here("til-r", "r-composition", "r_source_loc.csv"),
             col_names = c("language", "lines", "script")) |> 
  mutate(language = case_when(str_detect(language, "R,,") ~ "R",
                              str_detect(language, "C,,") ~ "C",
                              str_detect(language, "Fortran,,") ~ "Fortran"),
         lines = as.numeric(lines)) |> 
  distinct()

head(r_loc)

# A tibble: 6 × 3
  language lines script        
  <chr>    <dbl> <chr>         
1 R           20 .snow2.RR     
2 R            9 .multicore3.RR
3 R           15 .multicore2.RR
4 R           10 .multicore1.RR
5 R           25 .RSeed.R      
6 R           36 .Master.R     

Now, we can visualize the percentage of R Core sourcecode files by language using ggplot2:

library(ggplot2)
library(forcats)

r_loc |> 
  filter(!is.na(language)) |> 
  group_by(language) |> 
  summarise (n = n()) |> 
  mutate(rel.freq =  n / sum(n), accuracy = 0.1) |> 
  ggplot(aes(x = fct_reorder(language, desc(rel.freq)), y = rel.freq, fill = language)) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = scales::percent(rel.freq)),
    position = position_dodge(width = 0.9),
    vjust = -0.25,
    size = 4
  ) +
  theme_minimal() +
  labs(title = "Percentage of R Core Sourcecode Files by Language") +
  theme(plot.title = element_text(size = 14),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.x = element_text(size = 12),
        axis.text.y = element_blank()) +
  scale_fill_manual(values = c("R" = "#332288", 
                               "C" = "#882255", 
                               "Fortran" = "#44AA99"))

Or, we can visualize the percentage of R Core lines of code by language:

r_loc |> 
  filter(!is.na(language)) |> 
  group_by(language) %>% 
  summarise(sum_lines = sum(lines, na.rm = TRUE)) |> 
  ungroup() |> 
  mutate(percent = sum_lines/sum(sum_lines)) |> 
  ggplot(aes(x = fct_reorder(language, desc(percent)), y = percent, fill = language)) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = scales::percent(percent)),
    position = position_dodge(width = 0.9),
    vjust = -0.25,
    size = 4
  )+
  theme_minimal() +
  labs(title = "Percentage of R Core Lines of Code by Language") +
  theme(plot.title = element_text(size = 14),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.x = element_text(size = 12),
        axis.text.y = element_blank(),
        legend.position = "none") +
  scale_fill_manual(values = c("R" = "#332288", 
                               "C" = "#882255", 
                               "Fortran" = "#44AA99"))

It’s interesting to see how much goes into making R what it is: an ecosystem built on collaboration across languages and tools (which was the takeaway from the talk!). If you’re curious about R’s source code, give the script a shot!