MATLAB Central Newsreader - Beginner's loop

Beginner's loop

Matlab Beginner — Sun, 25 Oct 2009 20:48:29 -0400

hello everybody.

I am a beginner in matlab and I encounter some problems when
programming loops. I would like to understand what I do wrong and how
can I make it better. I hope there somebody can help me with this.

the problem is the following:

I need to write a function with branches that I need to use afterward
in computing some optimal values.
1. I wrote first a function file for the branching function:

function [c] = c1(x);

for i = 1:size(x)
    if x(i)>=0.5
        c(i) = (x(i)>=0.5)^2;
    else
        c(i) = x(i)-1;
    end
end

in principle I want it to go through all the elements of a vector that
I will give in later and evaluate it using the resp. expression.

2.Testing the function first: when In the main file I write:

x=linspace(-1,10,100)

c1(x)...

it evaluates only x(1)=-1... what am I doing wrong?

3. What I actually need is to use the function in this function file:
(e is a vector of variables e(1) and e(2))

function [firstordcond1, firstordcond2] = nest1seller(e,el,g,mu,b,Bet)

firstordcond2 = c1(e)-((Bet*b*el*(g-b)*((mu*g*e(1))./(mu*g*e(1)+(1-mu)
*b*e(2))-(mu*(1-g*e(1)))./(mu*(1-g*e(1))+(1-mu)*(1-b*e(2))))));
firstordcond1 = c1(2)-((Bet*g*el*(g-b)*((mu*g*e(1))./(mu*g*e(1)+(1-mu)
*b*e(2))-(mu*(1-g*e(1)))./(mu*(1-g*e(1))+(1-mu)*(1-b*e(2))))));
end

could anybody be so kind and help me with this thing?

many many thanks
a matlab beginner

Re: Beginner's loop

Phil Goddard — Mon, 26 Oct 2009 01:24:01 -0400

Given your input vector,
>> for i = 1:size(x)
is equivalent to
>> for i = 1:[1 100]
which is equivalent to
>> for i = 1:1

Hence you are only getting one pass through the loop.

Assuming x is always a vector you want
>> for i = 1:length(x)

Phil.

Re: Beginner's loop

NZTideMan — Mon, 26 Oct 2009 03:47:21 -0400

On Oct 26, 9:48 am, Matlab Beginner <matlabegin...@googlemail.com>
wrote:
> hello everybody.
>
> I am a beginner in matlab and I encounter some problems when
> programming loops. I would like to understand what I do wrong and how
> can I make it better. I hope there somebody can help me with this.
>
> the problem is the following:
>
> I need to write a function with branches that I need to use afterward
> in computing some optimal values.
> 1. I wrote first a function file for the branching function:
>
> function [c] = c1(x);
>
> for i = 1:size(x)
> if x(i)>=0.5
> c(i) = (x(i)>=0.5)^2;
> else
> c(i) = x(i)-1;
> end
> end
>
> in principle I want it to go through all the elements of a vector that
> I will give in later and evaluate it using the resp. expression.
>
> 2.Testing the function first: when In the main file I write:
>
> x=linspace(-1,10,100)
>
> c1(x)...
>
> it evaluates only x(1)=-1... what am I doing wrong?
>
> 3. What I actually need is to use the function in this function file:
> (e is a vector of variables e(1) and e(2))
>
> function [firstordcond1, firstordcond2] = nest1seller(e,el,g,mu,b,Bet)
>
> firstordcond2 = c1(e)-((Bet*b*el*(g-b)*((mu*g*e(1))./(mu*g*e(1)+(1-mu)
> *b*e(2))-(mu*(1-g*e(1)))./(mu*(1-g*e(1))+(1-mu)*(1-b*e(2))))));
> firstordcond1 = c1(2)-((Bet*g*el*(g-b)*((mu*g*e(1))./(mu*g*e(1)+(1-mu)
> *b*e(2))-(mu*(1-g*e(1)))./(mu*(1-g*e(1))+(1-mu)*(1-b*e(2))))));
> end
>
> could anybody be so kind and help me with this thing?
>
> many many thanks
> a matlab beginner

The great thing about Matlab is that expressions can be vectorised.
This means that for many cases, you don't need a loop at all.
For your case, you can do it like this:
c=x-1; % this creates a vector c with each element being x-1
c(x>=0.5)=(x>=0.5).^2; % Changes just the elements of c where x
>=0.5
Note: I have written this on the RHS:
(x>=0.5).^2;
because it is equivalent to what you have written, but it is nonsense.
Its value is always unity because x>=0.5 is logical 1 and when you
square it you get 1.
Maybe you meant:
(x-0.5).^2;
but who knows?

I glanced at your question 3, but its impenetrable. You'll have to
simplify it.