Beware with the mathematical bug in function kdelta(L1,L2) in itensor.
Maxima 5.9.0pre-cvs (with enhancements by W. Schelter).
Licensed under the GNU Public License (see file COPYING)
(C1) load(itensor);
;; Loading file /usr/share/maxima/5.9.0pre-cvs/share/tensor/itensor.fas ...
;; Loading of file /usr/share/maxima/5.9.0pre-cvs/share/tensor/itensor.fas is finished.
(D1) /usr/share/maxima/5.9.0pre-cvs/share/tensor/itensor.fas
(C2) kdelta([1,2,3],[A,B,C]);
(D2) - KDELTA([1], [A]) (KDELTA([2], [B]) KDELTA([3], [C])
- KDELTA([2], [C]) KDELTA([3], [B])) - KDELTA([2], [A])
(KDELTA([1], [C]) KDELTA([3], [B]) - KDELTA([1], [B]) KDELTA([3], [C]))
- (KDELTA([1], [B]) KDELTA([2], [C]) - KDELTA([1], [C]) KDELTA([2], [B]))
KDELTA([3], [A])
(C3) kdelta([A,B,C],[1,2,3]);
(D3) 0
(C4)
The output d3 is surely wrong. This function would give a correct one if the indices would be symbols.
There is no problem if the integer indices are in covariant position, result d2.
rgds,
Valerij.