I tried, and also could not get 0,
but this is due to the Schouten identity...
You can play with the Option Schouten of Contract (and also with
But it is non-trivial.
It really is better to just calculate scalar quantities.
I do not think that there are any bugs in DiracReduce. It is fairly simple code and biggest problem might be renaming of dummy indices,
but I think that works fine.
This archive was generated by hypermail 2b29 : 10/16/18-05:40:01 AM Z CEST