Hi
On Tue, Nov 29, 2022 at 11:11 AM Martin Juhás via AMBER
<amber.ambermd.org> wrote:
> possible FAILURE: check ired.vec.dif
>
> amber22/AmberTools/src/cpptraj/test/Test_IRED
>
> 43,60c43,60
>
> < 0.11900 0.02570 -0.02587 0.09724 0.15870 0.10523
> 0.05125
>
> ---
>
> > -0.11900 -0.02570 0.02587 -0.09724 -0.15870 -0.10523
> -0.05125
The issue here is just that the eigenvector sign has just been
flipped. In my experience whether this happens or not is dependent on
the underlying BLAS/LAPACK implementation and is not a cause for
concern (since it doesn't matter what "direction" the eigenvector
points). I've actually addressed this in the GitHub version of cpptraj
by using RMSIP to do eigenvector comparisons. You can safely ignore
this particular difference.
-Dan
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Nov 29 2022 - 09:30:02 PST
