I have a 6x6 adjacency matrix M. I want to work with only all the values 1 below the diagonal line (i.e.,tril(M,-1)). With the remaining values, I want to binarize the matrix such that the top 20% of values are converted to 1 and the lower 80% are converted to zeros. M is something like this:
-0.01 0.09 -0.16 -0.11 0.29 0.11
0.09 0.00 0.09 0.09 0.48 0.44
-0.16 0.09 0.01 -0.09 0.09 0.14
-0.11 0.09 -0.09 -0.01 -0.18 -0.04
0.29 0.48 0.09 -0.18 0.00 0.05
0.11 0.44 0.14 -0.04 0.05 0.00
Output would be:
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00
0.29 0.48 0.00 0.00 0.00 0.00
0.00 0.44 0.00 0.00 0.00 0.00
I was thinking I could sort all the values below the diagonal line then only pick the 20% but I'm not sure how to sort the entire matrix (as opposed to sorting a single column/row). Could anyone help?
Edit: I guess I could reshape M then sort it but that doesn't seem to be efficient. I would love to see a better approach.
You can try the following approach:
trilMask = logical(tril(ones(size(M)),-1)); %generates trilMask
M(~trilMask ) = 0; %zero-out values outsideMask
relevantVals = sort(M(trilMask)); %sort values
T = relevantVals(round(length(relevantVals )*0.8)+1); %calc threshold,
M(M<T) = 0 %perform thresholding
Result:
M =
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0.2900 0.4800 0 0 0 0
0 0.4400 0 0 0 0