Home

OTpy

OTpy is a fast Python package for models and algorithms in Optimality Theory (OT), alone with a simple easy-to-use user interface.

It reads files in a format illustrated in sample_input.txt and is able to generates legible report in HTML.

It implements Constraint Demotion (CD) and Fusional Reduction (FRed) for strict ranking model, and Generalized Iterative Scaling (GIS), Sequential Conditional GIS (SCGIS) and Nonlinear Conjugate Gradient (NCG) for Maximum Entropy model (Maxent).

The code is written in Python 3.3.3. The graphic interface is based on Tkinter within it. The generation of Hasse diagraph is supported by Graphviz and Python package pydot.

Quick start

• Download the repostory and run main.py using Python 3.3.

• Or download and run a stand-alone self-extracting executable from here.

• To generate Hasse diagram, you will need to install Graphviz. For people run with Python, pydot is also needed to operate .dot file.

Links

• Project home
• Stand-alone executables
• Poster: OTpy: A Tool for Constraint-based Linguistics

Acknowledgements

This is a summer project when I (Jinman Zhao) took part in 2014 UCLA Cross-disciplanary Scholars in Science and Technology (CSST) program. I would like to thank UCLA-CSST program for the chance and Prof. Bruce Hayes for his kind and inspiring advicing.