I think you can achieve this by a variation on run length encoding. First, encode the data based on condition 1, then use condition 2 to modify the decoding. (Note that I'm interpretting condition 2 to mean '**still** greater than 1' for three spots; let me know if this is correct).
% data
C=[0.3 0.8 0.3 1.3 0.6 1.1 3.1 5.1 3.1 0.5];
% make data nontrivial
C = [C, C]
% apply condition 1
L = C > 1
% get index of the start of each phase change
%
% note: the first index is always the start of the first phase, so
% we begin the mask with true
mask = [true, logical(diff(L))]
idx = find(mask)
% compute the run-length of each phase
%
% note: an 'extra' index is appended to the end of the data,
% (essentially a 'phantom' phase starting past the end of the data)
% in order to get the correct length of the last phase
rl = diff([idx, numel(C)+1])
% extract the values associated with the run-length encoding
v = L(mask)
% apply condition 2
% i.e. require all '1' phases to also have a run-length > 3
v = v & (rl > 3)
% decode the rle
D = v(cumsum(mask))
The input C and output D are (printed columnwise for easier comparison)
>> [num2str(C(:)), repmat(' ', numel(C), 1), num2str(D(:))]
ans =
0.3 0
0.8 0
0.3 0
1.3 0
0.6 0
1.1 1
3.1 1
5.1 1
3.1 1
0.5 0
0.3 0
0.8 0
0.3 0
1.3 0
0.6 0
1.1 1
3.1 1
5.1 1
3.1 1
0.5 0
Please accept this answer if it helps, or leave a note in the comments if I've missed something.
