ConIII is a Python package for solving maximum entropy problems with a focus on the pairwise maximum entropy model, also known as the inverse Ising problem. Support for Python 3.7 and higher only (v0.2.10 works with Python 2.7 but is no longer actively maintained).
If you use ConIII for your research, please cite the following:
Lee, E.D. and Daniels, B.C., 2019. Convenient Interface to Inverse Ising (ConIII): A Python 3 Package for Solving Ising-Type Maximum Entropy Models. Journal of Open Research Software, 7(1), p.3. DOI: http://doi.org/10.5334/jors.217.
This package is available on PyPI. It can be installed by running
$ pip install coniii
If you have trouble using pip
and PyPI, then you can always build this package from
source. Download the latest release from GitHub. Make sure that you are running Python
3.7 or higher. Inside "coniii", you must run
$ pip install .
Note: Using setuptools in the usual way of python setup.py install
will not work because
eggs are incompatible with cached jit functions.
If you would like to use the Enumerate
solver for system sizes greater than 9, you must
run enumerate.py to write those files yourself. This can be run from the install
directory. If you do not know where the installation directory is, you can find it by
starting a Python terminal and running
>>> import coniii
>>> coniii.__path__
Once inside the install directory, you can run in your bash shell
$ python enumerate.py [N]
where [N]
should be replaced by the size of the system. This will write the equations
for the Ising model in the {0,1} basis. On the other hand,
$ python enumerate.py [N] 1
specifies that the system should be written for the {-1,1} basis. For more details, see
the __main__
block at the end of the file enumerate.py.
A Jupyter notebook with a brief introduction and examples for how to use ConIII is available in the "ipynb" directory on the GitHub repository: [https://github.com/eltrompetero/coniii/blob/master/ipynb/usage_guide.ipynb]. The notebook is installed into your package directory if you used pip, or you can download it from the above GitHub link.
To use the notebook, install jupyter.
$ pip install jupyter
or if you are using the Conda package manager
$ conda install jupyter
Then, first copy the notebook file "usage_guide.ipynb" into a directory outside the "coniii" directory. Change to this directory and run
$ jupyter notebook
This should open the notebook in your default web browser.
In order to open the usage guide Jupyter notebook, you must have both Jupyter installed. To run the examples, you need a number of packages listed in "setup.py". These are all automatically installed into your Python path when you install ConIII through listed dependencies.
If you prefer to install the packages yourself, you can use the Python package pip. Open a terminal and run
$ pip install matplotlib 'multiprocess>=0.70.5,<1' matplotlib scipy 'numpy>=1.15.4,<2' 'numba>=0.43.0,<1' 'mpmath>=1.1.0' dill joblib
This package is only maintained for Python 3 as of v1.0.2 and has only been tested for Python 3.7.2. Check which version of Python you are running in your terminal with
$ python --version
ConIII has been tested on the following systems
- Debian 9 (Stretch)
- Mac OS X 10.12 (Sierra)
- Mac OS X 10.13 (High Sierra)
multiprocess module problems: n_cpus
kwarg can be set to 0 or 1 when the algorithm class
instance is declared. This will disable the parallel computing functionality.
Please file an issue on the GitHub if you have any problems or feature requests. Provide a stack trace or other information that would be helpful in debugging. For example, OS, system configuration details, and the results of unit tests. These can be run by navigating to the package directory and running
$ pytest -q
The package directory can be found by running inside python
>>> import coniii
>>> coniii.__path__
You may also need to install pytest.
$ pip install pytest
When updating to new 1.x.x versions please read the RELEASE_NOTES. There may be modifications to the interface including parameter names as we make future versions more user friendly.