Software for real-world models

[Uber Mensch]

Hello everybody

With the Model Thinking class in its finishing stages, is there any thought for a real-world model or a software that could help in modeling. This is a long term haul though and interested people can join the discussion.

I have just got a few ideas

• Simulation on agriculture
• State simulation
• Quant financing

[Rajesh Kasturirangan]

I strongly second this suggestion. Some of us - when we met at NIAS- wanted to build a systematic model building framework, not just for this course -  but for a future collective project. Just as there is the language R for statistics, we need another language M for modeling. Nishant, Sudhir and I have been thinking about this. Those of us who are interested in this project, please say aye.

Rajesh

--
Rajesh Kasturirangan
www.regularities.org

www.rajeshkasturirangan.com

[Uber Mensch]

I am in for it (in the long-term) and I am particularly interested in cellular automata and random walks and in running mass-scale simulations (though I am novice right now)

And I do have a collection of Python libraries(not created by me though) that could do the job.

But they need to refined so as to provide a framework.

A few of those libraires

• numpy
• scipy
• scikits package
• pytables
• numexpr
• theano
• networkx
• Pymc
• pandas

and more.

Once a few of us agree upon a common plan, we can start developing it and I am sure it could have a global participation once it has some serious utility value

[Rajesh Kasturirangan]

The common plan is the important part; we need to create a modeling framework that helps a lot of people. Its worth brainstorming a bit before we start adding libraries etc.

Rajesh

[Uber Mensch]

Spot on. And I agree with it. I am mostly free on Wednesdays and Saturdays from GMT 13:30:00 - 15:00:00. 

Would also try to fit in any time in weekdays GMT 12:30:00 - 15:00:00.

Also would suggest creating a separate user group at http://www.nabble.com/  for this brain-storming session and further development (this would enable better offline access) and we may also invite a few from the Model Thinking Course

Waiting for consensus to create a group there

[Ankur Pandey]

I've limited time and coding skills, but will keep an eye on this for any contribution I can make.

As of now, people with Python knowledge can check this out-  http://code.google.com/p/pyvote/ 

Regds,

-Ankur.

[Rajesh Kasturirangan]

I am out of the country now and will be back in May, but I suggest that those of us who have the time and desire to build a modeling framework send an email to this list (or a private email to me, if you prefer that) and we will go from there. 

Rajesh

[Ubermensch]

It seems that after the initial influx of a whole lot of neuroscientists, interest seems to have waned down in this group (this often happens). I have created a group for discussing about the framework here

http://mf.21204.n6.nabble.com/

and also created a inital post.

Would have a brainstorming session with interested people