How to manipulate arrays based on the transitions of the array elements?

I want o make every element of the existing array to 1, but whenever there is a transition like the element value changes from -1 to 1 or 1 to -1, I want to keep those elements as -1 -1, it will be more clear from example below:

If I have a array like this:

aa = [0 0 0 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 ]

and I want to manipulate it in such a way to get a new array like

bb = [1 1 1 1 1 1 -1 -1 1 1 -1 -1 1 1 1].

How should I do that?

Answer 1

[James Kristoff]

One way to do this is with a combination of logical indexing some knowledge of the problem. For instance we know that the transitions you want to mark with a -1 are located any time the difference of consecutive numbers is either 2 (1 - -1), or -2 (-1 - 1). Using that information I wrote the following code:

aa = [0 0 0 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 ];
bb = ones(size(aa));
bb(([abs(diff(aa)), 0] > 1 | [0, abs(diff(aa))] > 1)) = -1;

This code initiallizes a vector bb to be all ones, the same size as aa, then it uses a combination of the abs and diff functions to create a vector of logical (true or false) values that is true in the location that starts the transition and in the following location.

indx = ([abs(diff(aa)), 0] > 1 | [0, abs(diff(aa))] > 1);

Then it uses this logical to selectively modify bb, by assigning the value -1 to those locations.

[Micky]

Thanks James for the answer. I appreciate it!

Answer 2

[Micky]

Thanks James for the answer. I appreciate it!

Just for FYI. I also figured out another way i.e. Rotate the aa one time to the left and one time to the right by 1 and get the new array by multiplying the rotated arrays and finally replace the 0's by 1's.