Question

I am trying to create a sparse matrix which has a 2D pattern run down the diagonal. This is probably easiest to explain with a quick example.

Say my pattern is: [1,0,2,0,1]…

I want to create a sparse matrix:

    [[2,0,1,0,0,0,0...0],
     [0,2,0,1,0,0,0...0],
     [1,0,2,0,1,0,0...0],
     [0,1,0,2,0,1,0...0],
     [0,0,1,0,2,0,1...0],
     [...]]

The scipy.sparse.dia_matrix seems like a good candidate, however, I simply cannot figure out how to accomplish what I want from the documentation available. Thank you in advance

Answer 1

N = 10
diag = np.zeros(N) + 2
udiag = np.zeros(N) + 1
ldiag = np.zeros(N) + 1
mat = scipy.sparse.dia_matrix(([diag, udiag, ldiag], [0, 2, -2]), shape=(N, N))
print mat.todense()
[[ 2.  0.  1.  0.  0.  0.  0.  0.  0.  0.]
[ 0.  2.  0.  1.  0.  0.  0.  0.  0.  0.]
[ 1.  0.  2.  0.  1.  0.  0.  0.  0.  0.]
[ 0.  1.  0.  2.  0.  1.  0.  0.  0.  0.]
[ 0.  0.  1.  0.  2.  0.  1.  0.  0.  0.]
[ 0.  0.  0.  1.  0.  2.  0.  1.  0.  0.]
[ 0.  0.  0.  0.  1.  0.  2.  0.  1.  0.]
[ 0.  0.  0.  0.  0.  1.  0.  2.  0.  1.]
[ 0.  0.  0.  0.  0.  0.  1.  0.  2.  0.]
[ 0.  0.  0.  0.  0.  0.  0.  1.  0.  2.]]