Metadata-Version: 1.1
Name: wasserplan
Version: 0.1.3
Summary: 1-Wasserstein distances between districting plans
Home-page: https://github.com/vrdi/geometry-of-graph-partitions
Author: Parker J. Rule
Author-email: parker.rule@tufts.edu
License: UNKNOWN
Description: Wasserplan
        ==========
        
        ``wasserplan`` is a library for computing 1-Wasserstein distances
        between `GerryChain`_ partitions.
        
        Dependencies
        ------------
        
        ``wasserplan`` depends on GerryChain and `CVXPY`_. We recommend
        installing these dependencies with Anaconda (see the `official
        GerryChain installation instructions`_).
        
        Using
        -----
        
        We principally use this library to compute pairwise transport distances
        which we (approximately) project into 2D Euclidean space. A simple
        demonstration of this use case using Virginia congressional districts is
        available `as a notebook`_ `(open in Colab)`_.
        
        Performance
        -----------
        
        By default, ``wasserplan`` uses the open-source, lightweight `ECOS
        solver`_, which is included with CVXPY. However, this solver may exhibit
        performance issues for large graphs (thousands of nodes). Performance
        can be improved by using a commercial solver such as `MOSEK`_.
        
        License
        -------
        
        This library is available under the MIT License.
        
        .. _GerryChain: https://github.com/mggg/gerrychain
        .. _CVXPY: https://www.cvxpy.org/
        .. _official GerryChain installation instructions: https://github.com/mggg/gerrychain#installation
        .. _as a notebook: https://github.com/vrdi/geometry-of-graph-partitions/blob/master/wasserplan/VA%20ReCom%20demo.ipynb
        .. _(open in Colab): https://colab.research.google.com/github/vrdi/geometry-of-graph-partitions/blob/master/wasserplan/VA%20ReCom%20demo.ipynb
        .. _ECOS solver: https://github.com/embotech/ecos
        .. _MOSEK: https://www.mosek.com/
        
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Operating System :: OS Independent
Classifier: License :: OSI Approved :: BSD License
