Replace character in string/cell with the character above it

String/cell example: --> real string/cell is very long

a= { 'a' ; 'b' ; 'c' ; 'b'}

Wanted outcome:

a_new= { 'a' ; 'a' ; 'c' ; 'c'}

Answer 1

[the cyclist]

The general rule is not perfectly clear to me, from your small example. Do you mean that you want the even-numbered elements to be replaced? If so, then

anew = a;
a(2:2:end) = a(1:2:end-1);

If not, maybe you could explain in a little more detail.

[Dora de Jong]

Sorry I hope this will make it more clear.

String/cell example: --> real string/cell is very long

a= { 'a' ; 'b' ; 'c' ; 'b' ; 'h' ; 'g' ; 'b'}

Wanted outcome:

a_new= { 'a' ; 'a' ; 'c' ; 'c', 'h' , 'g' , 'g'}

Explanation outcome:

So every b wil be replaced with the character above it:

-first b by a

-second b by c

-third b by g

[the cyclist]

Here is one way:

a = { 'a' ; 'b' ; 'c' ; 'b' ; 'h' ; 'g' ; 'b'};

replace_idx = find(strcmp(a,'b'));
a_new = a;
a_new(replace_idx) = a_new(replace_idx-1);

This will not work if 'b' is the first element, so you will need to modify this algorithm in that case. But you haven't given info on what you want to do then.

[Dora de Jong]

Thank you so much!!

There is no b as first element, so this works perfect.

Answer 2

[Jan]

Maybe:

a_new = a;
a_new(2:2:end) = a(1:2:end);

Or:

index = repelemt(1:2:numel(a), 1, 2):
a_new = a(index)

[EDITED] after your comment: "So every b wil be replaced with the character above it"

index    = find(strcmp(a, 'b'));
a(index) = a(index - 1);

What about: {'a'; 'b'; 'b'} or {'b'} or {'b', 'a'}?

[Dora de Jong]

Thank you so much for your answers!!!

The last one works!