• Parallel Wordle Solver
  • Full Documentation
  • Build
  • Help
  • Default Behavior (no args)
  • Play 1 game at a time (-w)
  • Customizing Heuristics (-m)
  • Entropy Parallelization Params (-p, -f, -e)
  • Interactive Mode (-i)

Parallel Wordle Solver

Full Documentation

Please visit this pdf.

Build

Make sure to build the project before you run any of the commands below:

ᐅ stack build

Help

ᐅ stack exec wordleSolver-exe -- -h True +RTS -ls -N5
Usage: wordleSolver [OPTION...]
  -i Bool    --interactive=Bool  Interactive mode
  -m String  --method=String     Heuristic method
  -p String  --parMethod=String  Parallel method
  -f Int     --filterChunk=Int   Filter chunk size
  -e Int     --entropyChunk=Int  Entropy chunk size
  -w String  --word=String       Guess word
  -h         --help              Show help

Default Behavior (no args)

• Initiate 2,309 Wordle games and print out the sequence of turns it took to reach the final answer
• Utilizes the entropy heuristics
• Utilizes the "parMap" parallelization strategy
  • Utilizes filter_chunk_sz = 50 and entropy_chunk_sz = 30

ᐅ stack exec wordleSolver-exe -- +RTS -ls -N8
"cigar"[3]: ["soare","lucid","cigar"]
"rebut"[3]: ["soare","putid","rebut"]
"sissy"[3]: ["soare","flint","sissy"]
...

Play 1 game at a time (-w)

Instead of playing all 2,309 games, we can choose to play only one game by specifying the -w option. For example, below is an example of playing one round of Wordle with the answer "baker"

ᐅ stack exec wordleSolver-exe -- -w "baker" +RTS -ls -N5
["soare","clipt","gombo","deked","baker"]

Customizing Heuristics (-m)

While the default heuristics is Entropy, we can choose to use other heuristics. Here are three ways of solving the prove wordle game.

ᐅ stack exec wordleSolver-exe -- -m "frequency" -w "prove" +RTS -ls -N5
["arose","drone","trope","probe","prove"]
ᐅ stack exec wordleSolver-exe -- -m "minimax" -w "prove" +RTS -ls -N5  
["raise","drone","probe","prove"]
ᐅ stack exec wordleSolver-exe -- -m "entropy" -w "prove" +RTS -ls -N5  
["soare","pwned","prove"]

Entropy Parallelization Params (-p, -f, -e)

We can further customize the Entropy Heuristics for it to use different parallelization strategies:

• -p specifies the parallelization strategy. Users can choose between parList and parListChunk
• -f specifies the filter chunk size. It's an integer.
• -e specifies the entropy chunk size. It's an integer. Recall that by default, if a user specifies none of the above, we automatically execute -p parList -f 50 -e 30. Here is an example of using the parListChunk strategy with filter size of 64 and entropy size of 64:

ᐅ stack exec wordleSolver-exe -- -m "entropy" -p parListChunk -f 50 -e 50 -w "prove" +RTS -ls -N5  
["soare","pwned","prove"]

ᐅ stack exec wordleSolver-exe -- -m "entropy" -p parListChunk -f 50 -e 50  +RTS -ls -N5  
"cigar"[3]: ["soare","lucid","cigar"]
"rebut"[3]: ["soare","putid","rebut"]
"sissy"[3]: ["soare","flint","sissy"]
...

Interactive Mode (-i)

ᐅ stack exec wordleSolver-exe -- -i True +RTS -ls -N5
Please enter "soare" into NYT's Wordle game.
Enter NYT's response in 0/1/2 format:
...