help-octave
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GSoC: Student blogs


From: richard
Subject: Re: GSoC: Student blogs
Date: Fri, 25 May 2012 00:50:18 +0100
User-agent: Mutt/1.5.20 (2009-06-14)

On Thu, May 24, 2012 at 01:41:33PM +0200, c. wrote:  
> > I would need to ask 
> > about permissions before I can post a complete example for 
> > the model: and there are several variations; it's a work in 
> > progress. 
> 
> no need to show the full code, a stripped down example is often good enough

I suspected we might arrive here :-)

I will ask and I will think about a stripped down version. 

For information, the original development was in MathCad, and published here: 
http://www.debtdeflation.com/blogs/qed/

Note that the comments state the code is not Open Source.   

There are some other, related, sources on the internet. 

If you are content with a _very_ stripped down version add these : 
initial conditions: 
x4=0.15; y4=0.33; s4=5; m4=0;
Algebra: 
h=i;
g = (y4-m4)*exp((X(i,7)-x4)*s4/(y4-m4)) +m4; # phillips curve calculation
g=g+Variable; 

# This is perhaps one twentieth of the calculations carried out in the inner 
loop. 
# I'd suppose if you did this 20 times it might approximate to the 
# CPU load for the model. 

> I don't see your point here, what do you mean by "manipulating manually"?

Sometimes tools such as, quadratic optimisation, or least squared errors can 
be used to quickly maximise (or minimise) some feature of your model without 
much human intervention. For example, minimising the time taken for an 
aircraft to reach a particular altitude by hand would add little to my 
knowledge of how an aircraft works, hence it is best left to the computer 
to optimise the parameters in that example.  

This model is complex, in the sense that it is difficult to predict how 
it will behave - at least in my limited experience. Hence chosing which 
variable to alter, how to alter it etc, by manually changing some lines 
of octave code, helps me to understand the model. 
 
Richard



reply via email to

[Prev in Thread] Current Thread [Next in Thread]