Averaging Every nth column but the interval is different every time?

So I have coulmn vector of numbers 249 numbers long. The numbers go like this:

The problem is to average the numbers that are similar (delta = 1.5), i.e. points 1-8, 9-16, etc. The issue is that most of the points have 8 values, but some of them do not, so my previous code:

avg = mean(reshape(run,8,[]))';

does not work since the interval is 8, but actually the interval is not always 8 (example points 17-23 which is only 7 points). How would I go about averaging this data set at each of the points?

Answer 1

[Star Strider]

If your data do not have uniform ranges, a loop is likely the only option:

V = [ones(4,1)+rand(4,1)/50; ones(6,1)*4+rand(6,1)/50; ones(8,1)*8+rand(8,1)/50];   % Create Data

idx = [1; find(diff([0;V])>2); numel(V)];                                           % Index Changes
for k = 1:numel(idx)-1
    meanV(k) = mean(V(idx(k):idx(k+1)));                                            % Mean Of Index Range
end

It would be easier with your data, since the uniquetol function could be applicable, allowing the use of the accumarray function. Without your data, and seeing only the image of it you posted, it seems safest to use the find function and the loop.

Answer 2

[Matt J]

[~,~,G]=unique(floor(run));

avg=splitapply(@mean,run,G)