Prev: An exact 1-D integration challenge - 67 - (erf, sqrt, arcsin)
Next: governor parameter
From: Ashok on 4 Aug 2010 00:28
1. I read about the usage of rehash(). I use this function as soon as I finish executing the newly created *.m file. This seems to be taking the iterations further, although I have not completely verified it. Let me know your view on this usage.
2. I was not aware of any other linear equations solving tools available similar to solve(). I shall look it up and try to find that suits the requirement.
3. Yes, the objective function is differentiable. It is just of order 1.
From: Matt J on 4 Aug 2010 01:00
"Ashok " <hunt_ashok(a)yahoo.com> wrote in message <i3aq8k$rd7$1(a)fred.mathworks.com>...

> 2. I was not aware of any other linear equations solving tools available similar to solve(). I shall look it up and try to find that suits the requirement.
===============

Well, surely you knew that MATLAB can do things like the following

>> A=rand(3); x=[1;2;3]; y=A*x

y =

3.4448
2.9913
4.0956

>> xx=A\y %Invert the linear equation y=A*x

xx =

1.0000
2.0000
3.0000


> 3. Yes, the objective function is differentiable. It is just of order 1.
==================

Could be a problem. Except maybe for the active-set algorithm option of fmincon, all of the fmincon algorithms try to numerically approximate 2nd derivatives.
First  |  Prev  | 
Pages: 1 2
Prev: An exact 1-D integration challenge - 67 - (erf, sqrt, arcsin)
Next: governor parameter