[CPMD-list] Using SCALE with a triclinic crystal

[mkosmows]

Dear Juerg:

The coordinates generated with SCALE are in fact not correct.  Specifically, 
some of the x coordinates are too big.  Any idea as to why this is?

To get both sets of coordinates, I enter the literature coordinates into 
rpluto, pack a unit cell, remove the excess molecules thusly generated, and 
then save the resulting set of coordinates as either fractional or cartesian.


Thank you for your assistance,

Mark Kosmowski

>===== Original Message From Juerg Hutter <hutter at pci.unizh.ch> =====
>Hi
>
>have you checked that CPMD generates the correct
>cartesian coordinates from the scaled coordinates
>input?
>
>Juerg
>
>
>On Thu, 3 Apr 2003, mkosmows wrote:
>
>> Dear Juerg:
>>
>> I am using POINT GROUP AUTO with DELTA=.001
>>
>> When I input cartesian coordinates, I do not get the error warning.  I only
>> get the error warning when I put in fractional coordinates and the SCALE
>> keyword.
>>
>> My literature structure gives coordinates to 4 places (.xxxx) and I use 
rpluto
>> to pack the cell and save the coordinates of the cell, which continue to 
have
>> 4 places.  I then just cut and paste into the input file.
>>
>> I can send an input file if that would be helpful.
>>
>> Thank you,
>>
>> Mark Kosmowski
>>
>> >===== Original Message From Juerg Hutter <hutter at pci.unizh.ch> =====
>> >Hi
>> >
>> >you can use SCALE with the triclinic cell.
>> >The problem seems to be that you have specified
>> >a point group and the program couldn't verify that
>> >the input is compatible with this group.
>> >Have you specified the coordinates with enough
>> >accuracy? Are you sure that you have picked
>> >the correct point group, maybe you want to try also
>> >POINT GROUP
>> >AUTO
>> >
>> >Juerg
>> >
>> >On Wed, 2 Apr 2003, mkosmows wrote:
>> >
>> >> Dear CPMD community:
>> >>
>> >> I received the following message while doing a geometry
>> >> optimization:  MULTTB| SUPERCELL ATOMS FORM NO BASIS FOR THIS
>> >> POINT GROUP.
>> >>
>> >> I think this is because I tried to invoke the SCALE keyword
>> >> while using a triclinic crystal lattice.  Is there a way to
>> >> use fractional coordinates in crystals where alpha, beta and
>> >> gamma are not all 90 degrees?  Or does one simply need to
>> >> enter in cartesian coordinates in these cases?
>> >>
>> >> Thank you,
>> >>
>> >> Mark Kosmowski