This is related to the Mathematica package FeynCalc.

Is there a way to simplify the metric? I mean in the sense that when I evaluate something and I get a long expression involving lots of $$g^{12}$$ and other off diagonal entries, is there a way to let Mathematica know that I work in a flat space so that these off diagonal entries should be set to zero?

An example:



$$\mathrm{Tr}[\gamma^1\gamma^2\not{k}\not{p}] = 4(g^{12}k\cdot p+k^2p^1-k^1p^2)$$

where the indices might or might not be indices or powers. I don't know how to interpret that. I can only reside to dimension counting to guess which is which in larger expressions.

  • $\begingroup$ FeynCalc is actually not intended for working with explicit components of metric tensors and 4-vectors (unless you very well understand what you are doing), so whenever you get a result that contains such explicit components I would expect that there was a mistake in the input. It would be therefore very helpful if you could post the smallest working example of your code that leads to this kind of result. $\endgroup$
    – vsht
    Oct 12, 2015 at 13:28
  • $\begingroup$ @vsht OK I'll update the question. $\endgroup$ Oct 12, 2015 at 13:43
  • $\begingroup$ So with GA[1] and GA[2] you really mean particular Dirac matrices $$\gamma^1$$ and $$\gamma^2$$ right? $\endgroup$
    – vsht
    Oct 12, 2015 at 13:58
  • $\begingroup$ Yes that is correct. $\endgroup$ Oct 12, 2015 at 13:59

1 Answer 1


In principle, you can set the explicit values of the metric tensor by hand, e.g.

Inner[Set, Table[FCI@MT[i, j], {i, 0, 3}, {j, 0, 3}], {{1, 0, 0,0}, {0, -1, 0, 0}, {0, 0, -1, 0}, {0, 0, 0, -1}}];



returns only

$$ 4 (k^2 p^1 - k^1 p^2) $$

But since this is not the usual way FeynCalc is meant to be used, do not be surprised that for explicit indices many simplifications are simply not implemented.

  • $\begingroup$ Thanks for your answer. I think I wont touch anything. It's perhaps best if I rewrite my expressions in a more general form and simplify as much as I can by hand before plugging into FeynCalc. $\endgroup$ Oct 12, 2015 at 14:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.