https://github.com/healeycodes/markov-chain-generator
🤖 Build and consume Markov chain models to generate random realistic text.
https://github.com/healeycodes/markov-chain-generator
markov-chain markov-model text-generation
Last synced: 9 months ago
JSON representation
🤖 Build and consume Markov chain models to generate random realistic text.
- Host: GitHub
- URL: https://github.com/healeycodes/markov-chain-generator
- Owner: healeycodes
- License: mit
- Created: 2021-01-30T03:54:41.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2021-02-02T12:02:48.000Z (about 5 years ago)
- Last Synced: 2025-03-15T00:51:12.652Z (about 1 year ago)
- Topics: markov-chain, markov-model, text-generation
- Language: Python
- Homepage:
- Size: 123 KB
- Stars: 8
- Watchers: 2
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# markov-chain-generator
Use this library to generate random realistic text via Markov chain models.
See the example by running `python generate.py`.
## How?
Pass in a corpus, a state size, and a minimum length. Make sure the corpus uses capital letters and full-stops.
```python
from generate import get_source, build_model, generate_text
state_size = 2
min_length = 100
source = get_source('poe') # or some corpus as a string
poe = build_model(source, state_size)
text = generate_text(poe, state_size, min_length)
print(text)
'''
And really between two and three, there is no comfortable feeling. But the little spalpeen Mounseer Maiter-di-dauns that plumped his silf right down by the pond. Upon his own house, when wit and jollity reigned supreme-dampened a little, of course, expressed in regard to his conscience and himself. But although Hop-Frog, through the nose; the intestines through an avenue two miles long, and might have been a little amused. When we had great reason to congratulate ourselves upon our good fortune. Approaching the table, and in the way of indemnification inserted his left thumb in the external world, when, with many a fine thing.
'''
```
## Tests
Some basic tests can be ran with `python tests.py`
The Edgar Allen Poe and Oscar Wilde text corpuses were taken from https://github.com/nlp-compromise/nlp-corpus and I'm unsure of their license. The code is MIT.