ConSav - Consumption-Saving Models


A code library for solving and simulating consumption-saving models in Python using Numba JIT compiled functions.

Code repository

Notebook repository

It contains a mini-course on dynamic programming.

The consav package provides:

  • A Model class with predefined methods for e.g. saving and loading

  • A multi-linear interpolation module

  • Optimizers such as golden section search and newton-raphson

  • An upper envelope function for using the endogenous grid point method in non-convex models

  • Functions for interfacing easily with C++

All of the above is written to be Numba compatible.

The repository ConsumptionSavingNotebooks contains a number of examples on using the various tools and two models:

  • The canonical buffer-stock consumption model

  • A durable consumption models with non-convex adjustment costs