The goal of memorids is to provide a simple way to get good quality, memorable IDs.

R-CMD-check Codecov

For one reason or another I’ve implemented various versions of this over the years and wanted to make a simple, reusable package to do the work for me instead.

The IDs created are two words in English – an adjective followed by a noun – and are built from a random sample of a list of each.

memorids has enough adjectives and nouns to create over half a million unique combinations of ID.


You can install the development version from GitHub with:

# install.packages("devtools")


Let’s set a seed first so the results are reproducible:

memorids offers a single function, memorid(), which returns a random memorable ID.

#> [1] "memorable-salad"
#> [1] "unappetising-hall"
#> [1] "nonresistant-rain"

You can also change the separator.

memorids::memorid(sep = " ")
#> [1] "subliminal cuttlefish"


Please remember that memorids aren’t unique IDs. There are a finite number of them – 28,473,120 to be precise – and as a consequence, you should not use them in situations where a unique ID is required without also checking for collisions.

The safest way to do this is to track what memorids you’re currently using and check against that list for a collision when issuing a new memorid.

#> memorids info...
#>   Number of adjectives              : 17576 
#>   Number of nouns                   : 1620 
#>   Number of combinations            : 28473120 
#>   Max length of adjectives          : 23 
#>   Max length of nouns               : 21 
#>   Mean iterations before collision  : 7042

pre-commit hook

The repo contains a handy pre-commit hook to prevent committing when either data-raw/nouns.txt or data-raw/adjectives.txt is newer than R/sysdata.R.

To use it run the following in the main memorid directory:

ln -s ../../pre-commit.R .git/hooks/pre-commit