Thread: Integration of an expression, with matlab

Im having trouble understanding what is wrong with my program, it is intended to integrate the following expression (contained on the link) using the trapezium rule:

http://integrals.wolfram.com/index.jsp?expr=2*x*sin[P*Pi*x]^2&random=false

Where P=2
and it is bounded by the values pi and 0.

My program obtains the results 6.1295 when using my trapezium rule, and 4.6979 when it is done analytically.

It works for simple expressions i.e. sin (x) but for some reason it doesnt like this more complicated expression. I believe the error to be because of the 'P' in my expression but im unsure how to solve this, if this is the problem. Here is my program thus far:

n=100 %number of intervals
P=2 %constant
b=pi % upper bound
a=0 %lower bound

h=(b-a)/n;

f=inline('2*x*((sin(P*pi*x)).^2)')
i=1:1:n-1;

I=(h/2)*( f(a,P) + f(b,P) ) + h*sum(f(a+i*h,P))

%exact integration:
x=b;
Iexact01= - (cos(2*P*pi*x) + 2*P*pi*x*(sin(2*P*pi*x) - P*pi*x))/(4*(P^2)*(pi^2));
x=a;
Iexact02= - (cos(2*P*pi*x) + 2*P*pi*x*(sin(2*P*pi*x) - P*pi*x))/(4*(P^2)*(pi^2));
Iexact = Iexact01 - Iexact02

Any help would be greatly appreciated!!

Originally Posted by p123nky

Im having trouble understanding what is wrong with my program, it is intended to integrate the following expression (contained on the link) using the trapezium rule:

http://integrals.wolfram.com/index.jsp?expr=2*x*sin[P*Pi*x]^2&random=false

Where P=2
and it is bounded by the values pi and 0.

My program obtains the results 6.1295 when using my trapezium rule, and 4.6979 when it is done analytically.

It works for simple expressions i.e. sin (x) but for some reason it doesnt like this more complicated expression. I believe the error to be because of the 'P' in my expression but im unsure how to solve this, if this is the problem. Here is my program thus far:

Any help would be greatly appreciated!!

Try specifying the arguments in the inline statement:

f=inline('2*x*((sin(P*pi*x)).^2)','x','P')

just as it tell you in the help system, otherwise you don't know which argument it will decide is which.

CB

Thankyou for the reply, that seems to have done something. Ive tried it again though but it is telling me that my inner matrix dimensions don't agree, I'm rather confused by what this means. The error message is as follows:

f =

Inline function:
f(x,P) = 2*x*((sin(P*pi*x)).^2)

??? Error using ==> inlineeval at 15
Error in inline expression ==> 2*x*((sin(P*pi*x)).^2)
??? Error using ==> mtimes
Inner matrix dimensions must agree.

Error in ==> inline.subsref at 27
INLINE_OUT_ = inlineeval(INLINE_INPUTS_, INLINE_OBJ_.inputExpr, INLINE_OBJ_.expr);

using the m-file:

n=100 %number of intervals
P=2 %constant
b=pi % upper bound
a=0 %lower bound

h=(b-a)/n;

f=inline('2*x*((sin(P*pi*x)).^2)','x','P')
i=1:1:n-1;

I=(h/2)*( f(a,P) + f(b,P) ) + h*sum(f(a+i*h,P))

%exact integration:
x=b;
Iexact01= - (cos(2*P*pi*x) + 2*P*pi*x*(sin(2*P*pi*x) - P*pi*x))/(4*(P^2)*(pi^2));
x=a;
Iexact02= - (cos(2*P*pi*x) + 2*P*pi*x*(sin(2*P*pi*x) - P*pi*x))/(4*(P^2)*(pi^2));
Iexact = Iexact01 - Iexact02

Thankyou for having a look, it is really appreciated!!

Thankyou for your time, ive managed to figure it out. I failed to realise that simply typing '*' to multiply has confused the program.

Again, thankyou for your help!!

Originally Posted by p123nky

Thankyou for your time, ive managed to figure it out. I failed to realise that simply typing '*' to multiply has confused the program.

Again, thankyou for your help!!

I am trying to write a similar program but getting the same error message as the previous post. Please can you tell me what command you used instead of '*' to multiply? Thank you.

Originally Posted by jmparis

I am trying to write a similar program but getting the same error message as the previous post. Please can you tell me what command you used instead of '*' to multiply? Thank you.

Elementwise multiplication ".*"

CB