create cell array of individual years from a long time series

I have a matrix which is composed of a mixture of datevec and time series of measurements, e.g.

time = datevec(datenum('1966-01-01'):datenum('2009-12-03'));
data = rand(size(time,1),1);
Final = [time,data];

a = cell(length(unique(time(:,1))),1);

I would like to place all individual years in separate cells, I have created an empty cell array to house the values but am unsure about the most efficient way of moving the values from 'Final' to 'a'.

So far I have:

b = unique(Final(:,1));

for i = 1:length(a);
    a{i} = Final(find(Final(:,1)==b(i)),:);
end

This works fine but I was wondering if anyone had an alternative method?

Answer 1

[Andrei Bobrov]

try this is code (corrected)

time = datevec(datenum('1966-01-01'):datenum('2009-12-03'));
data = rand(size(time,1),1);
Final = [time,data];

[c c c] = unique(Final(:,1));
a2 = accumarray(c,Final(:,end),[],@(x){x});

[Richard]

Great suggestion, although I am unsure about how accumarray works in this manner. From the example, if I do not have the variable 'data' only the variable 'Final' i.e. one matrix with the dates and data, do I need to separate them into 2 variables and then follow the answer you have shown or can 'a2' be altered to use the matrix 'Final' instead of the vector 'data'?

[Andrei Bobrov]

corrected