cpm Roadmap

Performance improvement

We are always looking for ways to speed up the code. Any contribution aiming to do so is welcome.

Some specific areas we are looking to improve are:

  • Using numba to speed up the code and allow compilation of classes with @jitclass.
  • Reduce code duplication and improve code organization.
  • Improve the performance of the cpm.generators module to speed up simulations.
  • Improve shared memory usage and reduce the number of copies of the data.
  • Improve the data compilation process to speed up the majority of data cleaning happening in the modules during runs and exports.

Documentation

We would like to expand the documentation to include more examples and tutorials. We imagine that these will at some point migrate to their own repository, but for now, they will be part of the main documentation.

Testing

We would like to improve the testing coverage of the code.

New features

We are always looking for new features to add to the library. If you have an idea for a new feature, please open an issue to discuss it.

We already have a list of low-hanging fruit features that we would like to add:

  1. Expansion of the models classes. There are many model components out in the literature and we would like to add as many of them to the library as possible.
  2. Add model selection module, where we will incorporate model selection criteria and methods, which are also compatible with the hierarchical estimation approaches we already implemented.
  3. Add pre-built models to the library as ready-to-deploy applications.
  4. New and more efficitent Hierarchical and MCMC methods for model fitting (e.g. Variational Bayes Monte Carlo, Particle filtering).