bug?

[jjlvillarejo]

Hi,

I have resumed using xAct. Incredible performance of VarD!

Is this a bug?
---------------------
SSg = Sqrt[-Detg[]]*RicciScalarCD[]
SSg // RiemannToChristoffel // ChristoffelToMetric // ReshapeAction
ToCanonical[ContractMetric[VarD[g[a, b], PD][%]/Sqrt[-Detg[]]]]

-->
ToCanonical::cmods: Detected metric-incompatible derivatives {PD}.
Validate::inhom: Found inhomogeneous indices: {{-a, -b}, {s$219684, s
$219685}}.
-------------------------------------------------------
It does only happen with ContractMetric acting before ToCanonical, not
the other way around.

I am using xTensor 0.9.8 {2008,11,23} with the patch that was provided
in a previous thread (about Variational Derivative with xPert)

[RR]

Hi JJ,
I played around with your example a little bit, and now I guess the
problem is that in the expression

ContractMetric[VarD[g[a, b], PD][%]/Sqrt[-Detg[]]]

you have PD derivatives with upper indices. It seems that ToCanonical
gives errors when this happens. I don't think that this is a bug.

You can apply SeparateMetric[] before ToCanonical (since you apply
ContractMetric, you probably don't want this).
You can also give ContractMetric the option not to use upper
derivative indices
SetOptions[ContractMetric, AllowUpperDerivatives -> False];
e.g. in the beginning of your calculation.

Best,
RR

On Jun 9, 6:05 pm, jjlvillarejo <juan.jose.lopez.villar...@gmail.com>
wrote:

[jjlvillarejo]

Thank you, RR!

Certainly not a bug. Maybe could be set as a warning in ToCanonical?
(or in the documentation). I would say that a non-experienced user
does not expect to incur an error by using the combination ToCanonical
[ContractMetric[]] over any expresion. In my case, I have not been
using xAct for a long while, so it amounts to a similar situation! ;-)

Cheers,

JJ