You can update your example with the required dimensions, by making fake matrices with random values. For example, the below works and has the same dimensions as your original problem.
row_dim = 102
col_dim = 33
R = sparse(2:col_dim, 1:col_dim-1, 1, col_dim, col_dim)
size = row_dim * col_dim
O_density = 6 / size
F_density = 38 / size
O = sprand(row_dim, col_dim, O_density)
F = sprand(row_dim, col_dim, F_density)
for i = 1:100
O.nzval .+= i
F .= F * R .+ O
end
Note, I’m using density values based on the nonzeros values in your above examples.
What error are you getting?