Access elements of sparse matrix without find()

I am using a toolbox for algorithmic differentiation. The output is a sparse matrix like this:

(1,1) 0 (50,2) 0.17 (99,3) 0.456

How can I access these elements? With find() I lose the first element because one of the derivatives happens to be zero. The index of the first element is known so I assume there must be a way to do that.

Answer 1

[James Tursa]

If you store this matrix as sparse, then yes the (1,1) 0 element will not be physically stored and cannot be recovered with find. So you will need to store these elements differently. Maybe keep track of the derivative element indexing in another variable. I.e., you could have another sparse matrix that simply keeps track of where the derivatives are. e.g.,

(1,1) 1 (50,2) 1 (99,3) 1

Then this could be used to recover the index values for the derivatives, even the 0 ones.