Hi Laura,
Toric means there is no boundary when making connection. In 1d case, this means a group is considered as a ring and in 2d, a group is considered as a torus. Consider this small example:
>>> from dana import *
>>> G = np.zero(5)
>>> C = DenseConnection(G,G, np.ones(3), toric=False)
>>> print C.weights
[[ 1. 1. 0. 0. 0.]
[ 1. 1. 1. 0. 0.]
[ 0. 1. 1. 1. 0.]
[ 0. 0. 1. 1. 1.]
[ 0. 0. 0. 1. 1.]]
>>> C = DenseConnection(G,G, np.ones(3), toric=True)
>>> print C.weights
[[ 1. 1. 0. 0. 1.]
[ 1. 1. 1. 0. 0.]
[ 0. 1. 1. 1. 0.]
[ 0. 0. 1. 1. 1.]
[ 1. 0. 0. 1. 1.]]
You can see two extra 1 in the second case. The first correspond to the connection of G[0] with G[-1] and the second one to the connection of G[4] to G[5]. When toric is False, they are discarded since G[0] has not left neighbours and G[4] has no right neighbours. But in the toric case, they have.
The same is true for 2d but it is less convenient to write here.
I hope this answer your question.
Nicolas
> --
> You received this message because you are subscribed to the Google Groups "dana users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
dana-users+...@googlegroups.com.
> For more options, visit
https://groups.google.com/d/optout.