Group consecutive and non-consecutive values in a vector

Consider a vector V2 containing certain entries of some other larger vector V1 comprising nV1 entries.

How would one split up the vector V2 into segments comprising consecutive entries and output these segments as the cells of a cell array C?

Additionally, it could be the case that V2 'wraps', that is, if the first segment of C contains the value 1, and the last segment of C contains the value nV1, these two segments should be merged.

I would like to avoid dynamic re-sizing of vectors/arrays if possible.

Possible test case #1:

Input:

nV1 = 8;
V2 = [1];

Output:

C = {[1]};

Possible test case #2:

Input:

nV1 = 8;
V2 = [3,4,5];

Output:

C = {[3,4,5]};

Possible test case #3:

Input:

nV1 = 8;
V2 = [1,2,4,6,7];

Output:

C = {{[1,2]},{[4]},{[6,7]}};

Possible test case #4:

Input:

nV1 = 8;
V2 = [1,2,4,7,8];

Output:

C = {{[1,2,7,8]},{[4]}};

or

C = {{[4]},{[1,2,7,8]}};

[Jan]

I assume you do not need a nested cell

C = {{[4]},{[1,2,7,8]}}

but a cell is sufficient:

C = {[4], [1,2,7,8]}

Answer 1

[Jan]

nV1 = 8;
V2  = [1,2,4,6,7];

grp = cumsum([true, diff(V2)~=1]);
C   = splitapply(@(x) {x}, V2, grp);
if V2(end) == nV1
  C{1}   = [C{1}, C{end}];
  C(end) = [];
end
C

C = 1×3 cell array
    {[1 2]}    {[4]}    {[6 7]}

[Hyeongseok]

Nice solution! Thanks!

[Raymond MacNeil]

Brilliant, thanks.