Python package for solving assortative matching models with two-sided heterogeneity.
Python package for solving assortative matching models with two-sided heterogeneity. The theoretical framework behind the class of models solved by pyAM is described in Eeckhout and Kircher (2012).
$ pip install pyam
at a terminal prompt.
If you wish to contribute to the project you will likely want to install from source. First your will need to fork and then clone the source repository.
$ git clone https://github.com/YOUR-USERNAME/pyAM.git
Next create a new conda development environment
$ conda create -n pyam-dev python anaconda
activate the newly created development environment
$ source activate pyam-dev
and install additional dependencies not available within Anaconda.
$ pip install pycollocation $ pip install seaborn
Finally, change into your local clone of the pyam source directory and install the package in development mode.
$ pip install -e .
At the moment there are two example notebooks, one for positive assortative matching and one for negative assortative matching in the examples directory. The positive assortative matching works fine; the negative assortative matching, however, does not yet work (I suspect because of a poor algorithm for the initial guess).