Talk:Trace diagram

Needs work?

I hadn't heard of trace diagrams before, and was severely disappointed when I came here to learn about them. The article gives me an impression of what trace diagrams are and how they are used in computation, but they don't properly define how you compute using them, nor does it explain the notion of equivalence that tells you how it is used in proofs of multilinear identities. But maybe I am asking too much?

Some nitpicks: Trace diagrams are said to be directed graphs, but several examples shown are undirected. And the example for tr(A) isn't even a graph, since it appears to have no nodes at all, just an edge. It may be a perfectly good trace diagram even so – I have no idea. Also, the other two examples in the same figure have the letter C attached to them with no explanation. Ditto for the nodes marked with red asterisks. Hanche (talk) 22:27, 19 November 2009 (UTC)

Ciliation marks

Are the orange dots on the matrix diagrams the ciliation marks? If so, should the Gibbs-vector diagrams have them on the Levi-Civita dots? Without them they have a sign-ambiguity if I have understood correctly. --catslash (talk) 15:32, 17 February 2010 (UTC). I've just re-read it. It says (Ciliations of degree-3 vertices are omitted here because they do not change a diagram's function.), which I don't understand. --catslash (talk) 15:42, 17 February 2010 (UTC)

Yes, I clarified this to state it explicitly under the "drawing conventions" section. I also clarified the remark for 3-diagrams. This is not meant to be an exhaustive explanation, so I'd recommend checking out one of the references if you'd like to know more. Triathematician (talk) 12:36, 2 March 2010 (UTC)

Thanks for the drawing conventions clarification. I was being dense regarding the 3D Levi-Civita dots; of course rotating the vectors does not change the sign in an odd-number of dimensions. --catslash (talk) 13:22, 2 March 2010 (UTC)