Girth is a python package for estimating item response theory (IRT) parameters. In addition, synthetic IRT data generation is supported. Below is a list of available functions, for more information visit the GIRTH homepage.
Interested in Bayesian Models? Check out girth_mcmc. It provides markov chain and variational inference estimation methods.
Dichotomous Models
- Rasch Model
- Joint Maximum Likelihood
- Conditional Likelihood
- Marginal Maximum Likelihood
- One Parameter Logistic Models
- Joint Maximum Likelihood
- Marginal Maximum Likelihood
- Two Parameter Logistic Models
- Joint Maximum Likelihood
- Marginal Maximum Likelihood
- Mixed Expected A Prior / Marginal Maximum Likelihood
- Three Parameter Logistic Models
- Marginal Maximum Likelihood (No Optimization and Minimal Support)
Polytomous Models
- Graded Response Model
- Joint Maximum Likelihood
- Marginal Maximum Likelihood
- Mixed Expected A Prior / Marginal Maximum Likelihood
- Partial Credit Model
- Joint Maximum Likelihood
- Marginal Maximum Likelihood
- Graded Unfolding Model
- Marginal Maximum Likelihood
Ablity Estimation
- Dichotomous
- Marginal Likelihood Estimation
- Maximum a Posteriori Estimation
- Expected a Posteriori Estimation
- Polytomous
- Expected a Posteriori Estimation
Supported Synthetic Data Generation
- Rasch / 1PL Models Dichotomous Models
- 2 PL Dichotomous Models
- 3 PL Dichotomous Models
- Graded Response Model Polytomous
- Partial Credit Model Polytomous
- Graded Unfolding Model Polytomous
- Multidimensional Dichotomous Models
Via pip
pip install girth --upgrade
From Source
python setup.py install --prefix=path/to/your/installation
We recommend the anaconda environment which can be installed here
- Python 3.7
- Numpy
- Scipy
- Numba
import numpy as np
from girth import create_synthetic_irt_dichotomous
from girth import twopl_mml
# Create Synthetic Data
difficulty = np.linspace(-2.5, 2.5, 10)
discrimination = np.random.rand(10) + 0.5
theta = np.random.randn(500)
syn_data = create_synthetic_irt_dichotomous(difficulty, discrimination, theta)
# Solve for parameters
estimates = twopl_mml(syn_data)
# Unpack estimates
discrimination_estimates = estimates['Discrimination']
difficulty_estimates = estimates['Difficulty']
Without coverage.py module
nosetests testing/
With coverage.py module
nosetests --with-coverage --cover-package=girth testing/
Ryan Sanchez
[email protected]
MIT License
Copyright (c) 2020 Ryan Sanchez
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.