The wordle from stimim

wordle's Introduction

A program for the game Wordle.

words.json contains two lists:
- answers is the list of possible answers
- guesses contains other valid words that can be guesses
preprocess.py: preprocess words.json. The script will generate file database.json, which is required by all following programs.
algo1.py:

Implements an AI for the game. For each guess, the program finds the word whose possible results have the maximum entropy. For each valid guess, we define the entropy as:
```
n_i = | { answer | ComputeGuessResult(guess, answer) == result_i} |
N = SUM(n_i) == | { answer } |
entropy = -SUM(n_i / N * log(n_i / N))
```
python algo1.py <answer> to find the guessed words of a given answer.
benchmark.py: Computes number of required guesses for each answer.
manual.py: play the game interactively. When the program guesses a word, you need to input the result, it's a 5-character string, each character is either:
- o for letters in the word and in the correct spot.
- ! for letters in the word but in the wrong spot.
- (space) for letters not in the word.
For example:
```
guess 1:
raise
Please input the result:   !o   (this is "  !o ")
guess 2:
stowp
Please input the result: o      (this is "o    ")
guess 3:
sissy
Please input the result: ooooo  (this is "ooooo")
Done!
```

https://medium.com/@tglaiel/the-mathematically-optimal-first-guess-in-wordle-cbcb03c19b0a

The structure of my algorithm is based on this document. But, instead of using min-max to minimize the max size of remaining answers, my algorithm uses entropy instead.

Recommend Projects