Skip to content

Utilities for generating maximum entropy textures according to Victor and Conte 2012. **GitHub mirror**: development happens at https://gitlab.com/epiasini/metex

License

Notifications You must be signed in to change notification settings

epiasini/metex-github-mirror

Repository files navigation

Metex - Maximum Entropy TEXtures

License PyPI version Build status

Utilities for generating maximum entropy textures, according to Victor and Conte 2012.

Metex can be used as a standalone software from the command line, or as a Python package to generate and manipulate textures.

Requirements

  • Python 3
  • numpy ≥ 1.7
  • matplotlib

Installation

To install the latest release, run:

pip install metex

(depending on your system, you may need to use pip3 instead of pip in the command above).

This should install the metex package on your PYTHONPATH, as well as an executable script called metex to your regular PATH.

Testing

(requires setuptools). If metex is already installed on your system, look for the copy of the test_library.py and test_cli_interface.sh scripts installed alongside the rest of the metex files and execute it. For example:

python /usr/lib/python3.X/site-packages/metex/test/test_library.py
/bin/bash /usr/lib/python3.X/site-packages/metex/test/test_cli_interface.sh

Usage (command line)

Generate one sample of a fully random, square 10x10 texture, which will be saved to the current folder as '0.png':

metex 10

Save 10 samples of a 100x150 texture in folder named 'fig', with parameter 'alpha' set to -0.6. Name each sample 'fig_[n].png', where [n] takes on values 0,...,9:

metex --folder=fig --prefix=fig_ --n_samples=10 --alpha=-0.6 100 150

For more information on the command line parameters,

metex --help

(also see metex/core.py).

Usage (library)

The library implements a Texture class that represents the maximum entropy distribution of textures with a given level of a statistic and a given size. Objects of class Texture can be sampled, generating objects of class TextureSample. This is a subclass of numpy.ndarray with support for array-like manipulation, image generation for use in experiments (via matplotlib) and terminal-based pretty-printing (for messing around).

>>> import metex
>>> texture = metex.Texture(height=15, width=25, beta1=0.7)
>>> sample = texture.sample()
>>> print(sample)

⬛⬛⬛⬛⬛⬜⬛⬛⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜
⬜⬛⬛⬛⬛⬛⬛⬛⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜
⬛⬜⬛⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜
⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜
⬜⬜⬜⬜⬜⬜⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬛⬜
⬜⬜⬜⬜⬜⬜⬛⬛⬛⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜
⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬜⬜⬜⬜⬜⬜⬛⬛
⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜
⬜⬜⬜⬜⬜⬛⬜⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜
⬛⬜⬜⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬛⬜⬜
⬜⬜⬛⬜⬜⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬛⬜⬛⬛⬜⬜
⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬛⬛⬜⬜⬛⬛⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜
>>> print(sample[:5,:])

⬛⬛⬛⬛⬛⬜⬛⬛⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜
⬜⬛⬛⬛⬛⬛⬛⬛⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜⬜⬜⬜⬜⬜⬜⬜⬜
⬛⬜⬛⬜⬜⬜⬜⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬜⬜
⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛⬛
>>> print(sample[:,:5])

⬛⬛⬛⬛⬛
⬜⬛⬛⬛⬛
⬜⬜⬜⬜⬛
⬛⬜⬛⬜⬜
⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬜
⬜⬜⬜⬜⬜
⬜⬜⬜⬜⬜
⬛⬛⬛⬜⬜
⬜⬜⬜⬜⬜
⬜⬜⬜⬜⬜
⬛⬜⬜⬛⬛
⬜⬜⬛⬜⬜
⬛⬛⬛⬛⬛
⬜⬜⬜⬜⬛

The exact rendering of the texture will depend on your terminal.

Changelog

See the CHANGELOG.md file for a list of changes from older versions.

Authors

metex is maintained by Eugenio Piasini.

About

Utilities for generating maximum entropy textures according to Victor and Conte 2012. **GitHub mirror**: development happens at https://gitlab.com/epiasini/metex

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published