[ase-users] jacapo cell optimization broken

[Ask Hjorth Larsen]

Hi

On Thu, 25 Aug 2011, John Kitchin wrote:

> hmm... I see the problem, a StrainFilter is not an atoms object, and so has
> no arrays. It looks like at some point I added code to jacapo for
> determining if an atoms had changed. the reason for that is related to some
> issues with 32-bit and 64-bit representations of floats being just different
> enough to make atoms that were unchanged not equal the atoms read from a
> file that should have been the same.
> 
> I am not sure how to fix this without putting the (filtered) arrays onto the
> StrainFilter. any ideas?
> 
> John

You could implement the array access as a property ("getter") of 
StrainFilter which is actually a function call that retrieves a slice of 
the array from the atoms object.  If I understand the issue correctly.

Regards
Ask