Find all numeric values right after the NaN values in a column vector

Good evening,

I have a column vector such that

    vec = [1 ; 2 ; 3 ; 4 ; Nan ; 5 ; 6 ; 7 ; 8 ; Nan ; 9 ; 10 ; 11 ; 12 ; Nan];

I would like to find extract the elements of the vector ''vec'' that are located right after the NaN, such that I would get a new vector:

    new_vec=[5;9]

It is very easy to find the indices of all the NaN elements, however I don't know how to 'shift' those indices a place further to locate the values right after the NaN.

Thanks for your help in advance,

KMT

Answer 1

[Jan]

index  = isnan(vec);
result = vec([false; index(1:end-1)]);

Or in one line:

result = vec([false; isnan(vec(1:end-1))]);

This is "logical indexing" and "shifting" is simply to insert a FALSE at the beginning.

[Konstantinos Tsitsilonis]

Thanks for your answer it works perfectly, however I seemed to have run into another problem. In case my vector is

vec = [1 ; 2 ; 3 ; 4 ; nan ;nan;nan; 5 ; 6 ; 7 ; 8 ; nan ; 9 ; 10 ; 11 ; 12 ; nan]

Then this solution does not work. How could I generalise this for any number of consecutive ''nan'' at any point in the vector?

Answer 2

[James Tursa]

E.g.,

x = find(isnan(vec))+1;
x = x(x<=numel(vec));  % or x(x>numel(vec)) = [];
result = vec(x);