divide equal parts of a time series into a new matrix

Hi, its a simple question but I'm having problems with the conditions to impose. I have a column of scalonated values (like:1 2 3 1 2 3 1 2 3) along the time,as an input, with its respective responses in other column. Would like to separate each sequence of "123", to have the minimum value of the response of each sequence, and the respective time. I'd appreciate any help. Thanks

Answer 1

[Sean de Wolski]

So:

[val idx] = min(reshape(A,3,[]));  %each val is the respective minimum
idx = idx+(0:3:(3*length(idx)-1)); %original column index.

[Leonardo]

Thanks Sean, it works fine for finding the indexes at the end, but in the case of treating all 3 columns. In my case, I would like to have the minimum values of each sequence, just for the 3rd column. For example, if I have

T I O

1 1 3

2 2 8

3 3 6

4 1 5

5 2 7

6 3 2

7 1 3

8 2 5

9 3 2

want to have the minimum value of "O" for each sequence 123 of "I". Moreover, I know the limits of the interval of 123, but not the number of points.

Thanks in advance for your help.