All Projects → coolbutuseless → wordle

coolbutuseless / wordle

Licence: MIT license
Wordle helper for RStats

Programming Languages

r
7636 projects

wordle

R-CMD-check

The {wordle} package contains code to assist in finding good candidate words for Wordle.

“Wordle” itself is a guess-a-word puzzle playable online.

The game plays like the old ‘mastermind’ board game, but with letters instead of coloured pins. The gameplay is as follows:

  1. Enter a word as a guess for the hidden target word.
  2. Any letters which are within the hidden target word are coloured in yellow.
  3. Any letters which match exactly the letter in the hidden target word are coloured green
  4. Figure out a new candidate word as a guess for the hidden target word, and go back to Step 1.

In the following game of Wordle, the first guess was eaten, the second was arise, and then the third guess really only has one good option given the constraints revealed so far: aside. This was the hidden target word, which means the puzzle is solved!

The process of finding good candidate words given letters which have been seen so far is a good match for regular expressions. This package aims to help you find these good candidate words.

What’s in the box

  • play_wordle() Simple way to play a game of wordle in the R console
  • wordle_dict an ‘official’ list of words scraped from the Wordle website
  • WordleHelper R6 Class is the primary way of finding candidate words. It has the following methods:
    • $new() to start a new object to help with a new puzzle.
    • $words to get a list of all the remaining possible valid words given the words and responses that have been seen so far
    • $update() to notify the object of what the latest word was, and the colour responses received back from the game for each letter.
    • The WordleHelper class is a stateful wrapper around filter_words()

Advanced:

  • orthogonal_words sets of n words which encompass the top n*5 letters (by word frequency in the wordle_dict)
  • filter_words() is a stateless function for filtering a list of words by various constraints on letter position.
  • WordleGame R6 class represnting a wordle game engine
    • $new() to create a new game and internally choose a target word for this game
    • $try(word) to try a word and see what the respnse is in relation to the hidden target word
    • $share() create a block of unicode representing the evolution of the solution process.
    • play_wordle() is a simple helper function wrapping this class.

Installation

You can install from GitHub with:

# install.packages('remotes')
remotes::install_github('coolbutuseless/wordle')

Play a game of Wordle in your R console

wordle::play_wordle()

Help solve a puzzle with wordle::WordleHelper

In this example, after picking my favourite starting word, at each step I will just pick the first word in the alphabetical list of remaining possible words.

helper <- WordleHelper$new(nchar = 5)
length(helper$words)
#> [1] 12972
head(helper$words)
#> [1] "aahed" "aalii" "aargh" "aarti" "abaca" "abaci"

Initial word choice: arose

There are many opinions on a good starting word - I like: arose

Update puzzle state with the word played and the response:

helper$update("arose", c('grey', 'grey', 'grey', 'yellow', 'green'))
helper$words
#>  [1] "besee" "disme" "ensue" "esile" "fusee" "geste" "gusle" "issue" "istle"
#> [10] "lisle" "mesne" "piste" "pusle" "scene" "scute" "sedge" "segue" "seine"
#> [19] "seize" "selle" "semee" "semie" "sente" "shine" "shite" "shive" "shule"
#> [28] "shute" "sidhe" "sidle" "siege" "sieve" "since" "singe" "sithe" "sixte"
#> [37] "skene" "skite" "skive" "skyte" "slice" "slide" "slime" "slipe" "slive"
#> [46] "slype" "smeke" "smile" "smite" "snide" "snipe" "spice" "spide" "spike"
#> [55] "spile" "spine" "spite" "spule" "spume" "stede" "stele" "steme" "stile"
#> [64] "stime" "stipe" "stive" "stude" "stupe" "style" "styme" "styte" "suede"
#> [73] "suete" "suite" "sujee" "swede" "swile" "swine" "swipe" "swive" "sybbe"
#> [82] "sycee" "sythe" "teste" "unsee" "upsee" "usque" "visie" "visne"

Choose the first word: besee

Update puzzle state with the word played and the response:

helper$update("besee", c('grey', 'yellow', 'yellow', 'grey', 'green'))
helper$words
#>  [1] "esile" "scene" "siege" "sieve" "skene" "smeke" "stede" "stele" "steme"
#> [10] "suede" "suete" "sujee" "swede" "sycee"

Choose the first word: esile

Update puzzle state with the word played and the response:

helper$update("esile", c('yellow', 'yellow', 'yellow', 'grey', 'green'))
helper$words
#> [1] "siege" "sieve"

Choose the first word: siege

Success!

Orthogonal Word Sets

orthogonal_words are multiple lists of words from 1 to 5 words in a row. All words are drawn from wordle_dict.

Within each set of words there are no duplicated letters.

Within each set of words, the most common N letters from the wordle dictionary are represented.

E.g. The first 15 most common letters in the wordle dictionary are c("s", "e", "a", "o", "r", "i", "l", "t", "n", "u", "d", "y", "c", "p"). All the 3-word sets use each of these letters once (and once only) - no duplicated letters are allowed.

letter_freq[1:5]
#> [1] "s" "e" "a" "o" "r"
head(orthogonal_words[[1]])  
#>   word1
#> 1 soare
#> 2 aeros
#> 3 arose

letter_freq[1:10]
#>  [1] "s" "e" "a" "o" "r" "i" "l" "t" "n" "u"
head(orthogonal_words[[2]])  
#>   word1 word2
#> 1 sonar tuile
#> 2 suite loran
#> 3 sorel tuina
#> 4 roans tuile
#> 5 lores tuina
#> 6 soler tuina

letter_freq[1:15]
#>  [1] "s" "e" "a" "o" "r" "i" "l" "t" "n" "u" "d" "y" "c" "p" "m"
head(orthogonal_words[[3]])  
#>   word1 word2 word3
#> 1 carts poind muley
#> 2 coats pined murly
#> 3 cared moils punty
#> 4 cared ponty muils
#> 5 mares poind culty
#> 6 cures poind malty

letter_freq[1:20]
#>  [1] "s" "e" "a" "o" "r" "i" "l" "t" "n" "u" "d" "y" "c" "p" "m" "h" "g" "b" "k"
#> [20] "f"
head(orthogonal_words[[4]])
#>   word1 word2 word3 word4
#> 1 barks child pongy fumet
#> 2 barky child pongs fumet
#> 3 banks porgy child fumet
#> 4 porks chant bilgy fumed
#> 5 parky bongs child fumet
#> 6 porky bangs child fumet

letter_freq
#>  [1] "s" "e" "a" "o" "r" "i" "l" "t" "n" "u" "d" "y" "c" "p" "m" "h" "g" "b" "k"
#> [20] "f" "w" "v" "z" "j" "x" "q"
head(orthogonal_words[[5]])
#>   word1 word2 word3 word4 word5
#> 1 chunk gymps waltz fjord vibex
#> 2 glent brick jumpy waqfs vozhd
#> 3 gucks waltz fjord vibex nymph
#> 4 blunk cimex grypt waqfs vozhd
#> 5 glent prick jumby waqfs vozhd
#> 6 bling treck jumpy waqfs vozhd

Tweetable Wordle Game Engine

A playable game of Wordle in a tweet.

This was an exercise to see if I could simplify the WordleGame into just 280 characters. It’s mostly unreadable, and lacks safety checks, but it works!

#RStats #wordle in a tweet
s=\(x)el(strsplit(x,''))
t=s(sample(grep("^[a-z]{5}$",readLines('/usr/share/dict/words'),v=T),1))
while(1){g=s(readline("? "))
M=t==g
r=which(!M)
for(i in r)for(j in r)if(g[i]==t[j]){M[i]=2;r=r[r!=j]}
cat(paste0('\033[48;5;',c(249,46,226)[M+1],'m ',g))}

Expert Function: filter_words()

The WordleHelper R6 class is just a stateful wrapper around a core function called filter_words().

In general you wouldn’t need to call this function for solving a Wordle puzzle but it might come in handy for other word puzzles.

In this example, I’m searching for a word:

  • with 9 letters
  • starting with p
  • containing v and z somewhere, but not as the first letter
  • containing only one z
  • without an a or an o in it
words <- readLines("/usr/share/dict/words")

filter_words(
  words            = words,
  exact            = "p........",
  wrong_spot       = c("vz", "", "", "", "", "", "", "", ""),
  min_count        = c(v = 1),
  known_count      = c(z = 1, a = 0, o = 0)
)
#> [1] "pulverize"

Acknowledgements

  • R Core for developing and maintaining the language.
  • CRAN maintainers, for patiently shepherding packages onto CRAN and maintaining the repository
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].