Let’s use Wordle as an excuse to practice with string manipulation.

dict <- data.frame(word = read.csv("Collins Scrabble Words (2019).txt", 
                       header = FALSE, skip = 2)[, 1]) %>%
                     mutate(word=tolower(word))
sample_n(dict, 100) %>%
  unlist() %>%
  unname()
##   [1] "afferent"        "waybill"         "wammul"          "mitigates"      
##   [5] "castlings"       "socialises"      "dawding"         "prolocutors"    
##   [9] "brandishes"      "prehending"      "coleopters"      "inthroning"     
##  [13] "embryologists"   "riotous"         "misspent"        "consultantships"
##  [17] "paramounts"      "tabogganing"     "retypes"         "polarograph"    
##  [21] "shivery"         "diandries"       "inegalitarian"   "teuchats"       
##  [25] "easts"           "creates"         "pedanticisms"    "semiologists"   
##  [29] "certificates"    "oratrixes"       "utterless"       "nilghau"        
##  [33] "superpersons"    "scowrie"         "calisthenics"    "caricatured"    
##  [37] "hornists"        "voluntary"       "orogenic"        "clairaudience"  
##  [41] "figurately"      "cataract"        "mopoke"          "gooks"          
##  [45] "scalableness"    "vinegary"        "humidities"      "monochlorides"  
##  [49] "outacting"       "unthatch"        "endangerer"      "pollman"        
##  [53] "recertify"       "notching"        "virandas"        "verberations"   
##  [57] "forevermore"     "microcopy"       "pabulums"        "hearsy"         
##  [61] "emperorships"    "blindfolded"     "piscator"        "magnificat"     
##  [65] "vacantnesses"    "reassessed"      "scrimpers"       "gizzards"       
##  [69] "rearhorses"      "marjoram"        "surmasters"      "octostyle"      
##  [73] "smoothpate"      "sokol"           "conjunctions"    "rizzoring"      
##  [77] "dysphemistic"    "douc"            "frontlists"      "caviars"        
##  [81] "civils"          "clysters"        "neurular"        "midstories"     
##  [85] "endogenous"      "fourfoldness"    "thistledowns"    "methanations"   
##  [89] "syndicate"       "ignorances"      "satsangs"        "jacklights"     
##  [93] "whither"         "slushily"        "comices"         "pileorhizas"    
##  [97] "spivvery"        "weblish"         "hexaemeric"      "swealings"

January 18, 2022

First word ‘adieu’, all grey!

⬜⬜⬜⬜⬜

Let’s filter out all those vowels and the number of characters.

out <- dict %>%
  filter(str_detect(word, "a|e|i|u", negate=T),
         nchar(word)==5)
nrow(out)
## [1] 1200

1,200 words to work with

sample_n(out, 75) %>%
  unlist() %>%
  unname()
##  [1] "cocco" "proto" "jomon" "hymns" "kloof" "donny" "bohos" "motts" "poboy"
## [10] "poots" "boozy" "phons" "orlop" "poynt" "myops" "typos" "kroon" "vlogs"
## [19] "trons" "cobbs" "sodom" "flogs" "poopy" "golly" "sybow" "sooky" "molds"
## [28] "crook" "octyl" "torts" "phots" "bowrs" "wormy" "jowls" "dolor" "gotch"
## [37] "trots" "hoosh" "corno" "socko" "toffy" "comby" "jolty" "goory" "swoon"
## [46] "clomp" "polls" "proyn" "mosks" "smogs" "sprod" "forth" "jonty" "noops"
## [55] "nolls" "shows" "north" "wooly" "showd" "hoppy" "glost" "jolls" "bonks"
## [64] "zocco" "bobby" "bowls" "looms" "goopy" "cyton" "sords" "sowff" "konks"
## [73] "prosy" "snobs" "crool"

Second try ‘strop’. R, o, and p yellow

⬜⬜🟨🟨🟨

Let’s filter for most common consonants, as well as the information from the previous guess.

out2 <- out %>%
  filter(str_detect(word, "t|n|s"),
         str_detect(word, "r|o|p"))

nrow(out2)
## [1] 909

909 words to work with1

sample_n(out2, 75) %>%
  unlist() %>%
  unname()
##  [1] "molds" "tophs" "tronc" "coopt" "wonts" "confs" "torcs" "zobos" "wonky"
## [10] "cloys" "omovs" "gonof" "shogs" "frory" "moots" "plong" "scrod" "corso"
## [19] "jolls" "tofts" "shock" "bosky" "poynt" "whort" "forky" "gyron" "dorts"
## [28] "gymps" "nobby" "thoft" "stong" "snobs" "volks" "yorks" "towzy" "flogs"
## [37] "rocks" "yolks" "shorn" "shool" "tromp" "showy" "troth" "gongs" "forgo"
## [46] "scoop" "pooks" "shots" "pomps" "hoods" "softs" "proso" "worms" "boons"
## [55] "story" "flops" "sycon" "flows" "prost" "props" "koffs" "hoghs" "torot"
## [64] "holon" "stonk" "cowks" "sownd" "posho" "rotls" "rooms" "yocks" "notch"
## [73] "bolos" "boxty" "lownd"

Third try ‘poncy’. p and y green, o yellow.

🟩🟨⬜⬜🟩

In fact that’s a regression since I did not choose a word with r or o.

Let’s filter for the first and last letters.

out3 <- out2 %>%
  filter(substr(word, 1,1)=="p",
         substr(word,5,5)=="y")

nrow(out3)
## [1] 14

Only 14 words possible.

out3 %>%
  unlist() %>%
  unname()
##  [1] "phony" "poncy" "pongy" "ponty" "popsy" "porgy" "porky" "porny" "porty"
## [10] "potsy" "potty" "powny" "prosy" "proxy"

Next try pygmy, no help.

🟩⬜⬜⬜🟩

In fact that’s an even worse regression because I forgot to use a word that included o and r again. Let’s filter for o and r so I won’t forget, and also remove the consonants we’ve used so far.

out4 <- out3 %>%
  filter(str_detect(word, "o|r"),
         str_detect(word, "d|s|t|n|c|g|m", negate=T))
nrow(out4)
## [1] 2

Only two words left.

out4 %>%
  unlist() %>%
  unname()
## [1] "porky" "proxy"

Tried proxy on fifth attempt, got it!

Zero credit for using any vocabulary skills (I still don’t know what strop or poncy mean). But it’s a good occasion to do some string manipulation. To keep some skill involved, next time I won’t filter for most common consonants, but only filter out based on past guesses.

Wordle 213 5/6

⬜⬜⬜⬜⬜ ⬜⬜🟨🟨🟨 🟩🟨⬜⬜🟩 🟩⬜⬜⬜🟩 🟩🟩🟩🟩🟩

  1. I later noticed that the sequential calls to str_detect did some additional filtering↩︎