Is there a way to extract dynamic values of timeseries without hardcoding?

I have a multiple timeseries that goe as follows:

Subject 1 (13x1):

1
1
1
2
2
2
5
5
5
5
8
8
8

Subject 2:

1
1
1
4
4
4
4
4
7
7
9
9
9
9
9
9

I want to calculate the difference between every change and the first value (1), and then I also want to caclualte the difference between the multiple conditions (e.g. 7-4=3, 5-2=3, etc). For one subject, I could hard code the section of the timeseries as follows and do as follows:

Example for subject 1:

timeseries(11)-timeseries(2)=6

Example for subject 2:

timeseries(4)-timeseries(1)=3

Is there a way to do this without hardcoding the part of the timeseries?

Thanks a ton

[Matt]

The function unique (https://fr.mathworks.com/help/matlab/ref/double.unique.html) can help you do that.

C = unique(timeserie);
C-timeserie(1)% this is the difference between all the multiples elements and the first element

if you want the difference between all the uniques elements you can do a loop on the elements of C to compute all the possible variations. Something like this :

for ii=1:length(C)
    for jj=1:length(C)
    diff(ii,jj) = C(ii)-C(jj);
    end
end

[nines]

thank you!!!!