I want to create a n by 1 matrix. The FDF2=2 is on the top row, while BDF2=3 is on the bottom. And the values of the middle row should all equal to 3. How do I input the values in such matrix?

Should I create a zeros(n+1,1) matrix first?

Answer 1

[John D'Errico]

I'm not sure why you would initially create a VECTOR of length n+1, if your final goal is an nx1 vector. :)

No. There is no need to preallocate the vector with zeros, although your mentioning that idea is the only reason I was willing to answer your question. There are lots of ways to create this vector though. A simple one is:

V = [FDF2;repmat(3,n-2,1);BDF2];

An alternative is to preallocate a vector, then stuff the first and last elements.

V = ones(n,1)*3;
V([1 end]) = [FDF2,BDF2];

Lots of other ways to do this, but that covers the first ones I might think of. You can see that the first solution involves concatenation of the indicated elements, whereas the second had me create a vector of the final size, then stuffing the first and last elements as desired. Either is fine.

[kingsley]

got it! thank you. I'm new to mathlab. So the question I ask might sound stupid. But what if the value on the on the middle rows. And it has to be 2*fi, where i is from 1 to n-1, and the function of f=sin(i)? Do I need to write a loop in that case ?

[John D'Errico]

Learn to think in terms of vectors. What is

1:(n-1)

This is a row vector. But if you want a column vector, then what does this do?

(1:(n-1))'

Can you compute the sine of ALL of those values in one call to sin? Of course you can.

Think at a high level, in terms of vectors & arrays, not in terms of loops.