Thread Subject:

shifting vectors according to an input integer

I'm trying to write this function that will shift whatever input vector the amount of times specified. So there are two inputs and one output.
An example:
input 1: [1 2 3 4 5]
input 2: 4 (could be a negative integer as well)
So for the above case, the function would operate like this:
[1 2 3 4 5]
[5 1 2 3 4]
[4 5 1 2 3]
[3 4 5 1 2]
[2 3 4 5 1]
It has shifted 4 (input 2) times and the desired answer is [2 3 4 5 1].
If the number were negative:
input 1: [1 2 3 4 5]
input 2: -2

[1 2 3 4 5]
[2 3 4 5 1]
[3 4 5 1 2]
It would go backwards. This function has to work for columns as well, however they are all single rows or single columns.
My work so far:
 B=1:length(vec); %vec is input 1
mod(shift, length(vec)) %shift is input 2
So it has something to do with the mod function...because that's what we're practicing with. Now I'm pretty much stumped how to proceed?

Thank you

Have you seen CIRCSHIFT?

"Matt Fig" <spamanon@yahoo.com> wrote in message <i6cagl$7ns$1@fred.mathworks.com>...
> Have you seen CIRCSHIFT?

Not allowed to use that unfortunately. Only using mod, :, vec(vec())...etc.

"Raymond " <atlantaswagsurfers3@yahoo.com> wrote in message <i6cb3e$dmo$1@fred.mathworks.com>...
> "Matt Fig" <spamanon@yahoo.com> wrote in message <i6cagl$7ns$1@fred.mathworks.com>...
> > Have you seen CIRCSHIFT?
>
> Not allowed to use that unfortunately. Only using mod, :, vec(vec())...etc.

Hi

first guess:
B=1:numel(vec)
e.g. B=[1 2 3 4 5]

take your B, add the SHIFT
suppose SHIFT=2
C=B+SHIFT
e.g. C=[3 4 5 6 7]

now the 3 4 5 are fine, but the 6 7 are wrong, they should be 1 2
so try using mod here
C=mod(B+SHIFT,numel(vec));

I know this isn't right (and yes I do have a solution), but maybe you can figure out what to do next?

Ross

"Raymond " <atlantaswagsurfers3@yahoo.com> wrote in message <i6cb3e$dmo$1@fred.mathworks.com>...
> "Matt Fig" <spamanon@yahoo.com> wrote in message <i6cagl$7ns$1@fred.mathworks.com>...
> > Have you seen CIRCSHIFT?
>
> Not allowed to use that unfortunately. Only using mod, :, vec(vec())...etc.

Like this?
%Vector to shift and shift
A = [1 2 3 4 5];
shift = 3;

%Matrix of combos (twice the size for the purpose of negative shifts)
T = repmat(triu(toeplitz(A))+tril(toeplitz([0 A(end:-1:2)])),2,1);

%Row Selection
T(max(rem(shift,5)+1,numel(A)+abs(mod(shift,5))+1),:)

:)