Question

I’m eager to know,

• how many package names on CRAN have two, three, N characters?
• which combinations have not yet been used (“unpoppler”)
• how many package names use full-caps, or camelCase?
• how many package names end in 2?

I think it might reveal some interesting facts.

Edit: bonus points for animated graphics showing the time-evolution of CRAN packages.

Answer 1

A better way than scraping a web page to get the names of packages is to use the available.packages() function and process those results. available.packages() returns a matrix contains details of all packages available (but is filtered by default — see the Details section of ?available.packages for more).

pkgs <- available.packages(filters = "duplicates")
nameCount <- unname(nchar(pkgs[, "Package"]))
table(nameCount)

> table(nameCount)
nameCount
  2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17  18  19  20  21 
 32 311 374 360 434 445 368 277 199 132  99  56  56  43  22  19  18   2  12   8 
 22  24  25  31 
  5   2   1   1

Using nameCount we can select packages with names containing any number of characters without needing to resort to regexp etc:

> unname(pkgs[which(nameCount == 2), "Package"])
 [1] "BB" "bs" "ca" "cg" "dr" "ez" "FD" "ff" "HH" "HI" "iv" "JM" "ks" "M3" "mi"
[16] "np" "oc" "oz" "PK" "PP" "qp" "QT" "RC" "rv" "Rz" "sm" "sn" "sp" "st" "SV"
[31] "tm" "wq"