Algorithms / routines for possible inclusion in QuantEcon.py or QuantEcon.jl


#1

This is topic for informal discussion of the kinds of routines or algorithms we might like to see implemented in one or both of the QuantEcon libraries over time. They can be ambitious or minor, vague or specific.

Here’s a start:

  • Game theory and particularly computation of Nash Equilibria (similar to or perhaps using) Gambit
  • A set of routines for solving asset pricing models (I have some ideas for this I’ll get to soon)
  • Matching algorithms, such as these (stable marriage problem)
  • abreu pearce stacchetti for repeated games (on the way, I believe)
  • more repeated games stuff

Not projects I have much knowledge of, but here are some other econ related software projects:


#2

Author: Ken B

I’d be happy to help out if you want to include some Agent Based Models in QuantEcon. I’ve been writing some approaches in Julia and plan to do more in the future.

– Ken


#3

Hi Ken,

Thanks for your response and we’d be happy to hear more.

One thing I should mention: The basic rule for the libraries is that they are for implementing algorithms or routines that have multiple applications across different models or use cases. For more specific code that is tied to a given model (but still interesting and potentially useful), please consider submitting to the notebook gallery.


#4

Author: oyama....@gmail.com

For experimental economics, oTree looks really great.
One of the nice things is that it is open source (written in Python).

Matching algorithms are easy to implement (because the algorithms are already there).
I have some code written here (for marriage and college admission problems).
The difficult part is to decide on the user interface, the data structure for the input and the output.


#5

Author: David Pugh

I think it would be useful to have more routines for handling continuous time. I have already contributed an IVP module and have written a BVP solver that can be included if there is interest (following a little cleaning and adding some documentation). Solving systems of partial differential equations with specific applications in heterogenous agent macro (i.e., work of Ben Moll and others) as well as finance. There exist a number of excellent Python libraries for solving physical systems of PDEs however it is not clear to me whether or not these would be suitable for economic models.


#6

Auction related ideas:


#7

BLP Algorithms


#8

Here is a collection of software tools that may be of interest and could contain a lot of potential future projects (interface to python/julia etc.):

http://www2.econ.iastate.edu/tesfatsi/acecode.htm


#9

Suggestion by Simon Hoof:

  • Projection methods for solving for value functions, similar to the methods in the CompEcon toolkit by Miranda and Fackler.

This could be a topic suited either to an independent notebook or to one of the libraries. I would like to see this done.


#10

Code related to mechanism design: