Markov chain-based name generator based on Sam Twidale implementation in Haxe. consider this work to be a python port with functionnally the same level of functionality. Feed it a list of training data, and it'll spew out a bunch of new words.
It comes with the training data compiled by Sam, and made available under
Creative Commons BY-NC 4.0.
If you appreciate this work, consider buying me a coffee.
Install using:
pip install markovname
This software has been developped and tested with python 3.9. Any earlier version might not work.
A minor dependency on wcwidth, make sure it is installed using pip, your
package manager or whatever method you are familiar with. It is only necessary
if you plan to use the main module. If you're only interested in the generator,
then wcwidth can be ignored.
If you have poetry available, you can initialize a virtual environment with:
poetry install
The software reads sample data from json formatted datasets. See the data folder of this repository for examples.
usage: __main__.py [-h] [--list] [-n NUMBER] [-o ORDER] [-p PRIOR] [-g KEY] [-V] DATA
positional arguments:
DATA Path containing the data sets.
optional arguments:
-h, --help show this help message and exit
--list List the various dataset available to the software.
(default: False)
-n NUMBER, --number NUMBER
Amount of names to generate. (default: 1)
-o ORDER, --order ORDER
Highest order of model to use. Will use Katz's back-
off model. It looks for the next letter based on the
last "n" letters. (default: 3)
-p PRIOR, --prior PRIOR
The prior adds a constant probability that a random
letter is picked from the alphabet when generating a
new letter. Must be a number between 0 and 1.
(default: 0)
-g KEY, --generate KEY
Which dataset to use. If not specified, use a random
available set instead. (default: None)
-V, --version show program's version number and exit