I have the following equation: $$f_i^´(t+1)=\sum_{jk}R_{i|jk}\tilde{f_j}(t)\tilde{f_k}(t)$$
It is about evolution of a population.
I use this equation in my python program in the following way:
re = numpy.zeros((l, l, l))
wp = numpy.zeros((l,2))
#some code with edits the entries of re and wp
for x in range(N):
wp[:, 0] = numpy.tensordot(numpy.tensordot(re, wp[:, 0], axes=1), wp[:, 0],
axes=1)
Now I am not interested in how every time step looks like but more in how the fixed points look like. Thus I want to know the outcome for large N(=after many time steps). If this equation would have a Matrix instead of a Tensor, then I could take the Matrix to the power of N, save this matrix and use it for all different kind of vectors to get their final state. This is of course much faster. Now my question is, if something like this is also possible for a rank 3 Tensor which I have here? Then my program could be much faster. I am using Tensors for the first time so I hope this question isn't too stupid.