Ahhh!
When you run:
cluster_tot{:,3}
It extracts every cell in that column. Each cell is then passed into max() as it's own variable. So for example:
x = num2cell(rand(5,1))
max(x{:})
Is equivalent to
max(x{1},x{2},x{3},x{4},x{5})
Max doesn't work like this. If the cells are all the same size, you should be able to horizontally or vertically concatenate them in order to take the max:
max([cluster_tot{:,3}])
Which would first combine them into one input and then take the max of it.
If they're not the same size, you'll need to use cellfun first.
max(cellfun(@(x)max(x(:)),cluster_tot(:,3)))
Note the use of () instead of {} because I want cellfun to operate on the cells not the contents in them.
