error in pw2bgw?

Submitted by rangel on Mon, 02/10/2014 - 11:36

Forums 

Bug reports

Hi,
I found a problem in the code when omitting celldm in espresso.

To illustrate this:
Please consider the silicon example in the BGW suite:

In 02-wfn:
Instead of using:
celldm(1) = 10.2612
and
CELL_PARAMETERS
0.000000000 0.500000000 0.500000000
0.500000000 0.000000000 0.500000000
0.500000000 0.500000000 0.000000000

remove "celldm(1)= 10.2612" and change the cell parameters to:
CELL_PARAMETERS bohr
0.000000000000 5.130600000000 5.130600000000
5.130600000000 0.000000000000 5.130600000000
5.130600000000 5.130600000000 0.000000000000

This is still correct for espresso. I guess it reads the lattice constant from other files, since it is correct in the output file.
But BGW will fail at the epsilon step. It complains that one k-point is not found..

Best
Tonatiuh

Submitted by rangel on Mon, 02/10/2014 - 13:49

I think it might be that pw2bgw expects the CELL_PARAMETERS to be given in units of the lattice parameter..
I am just guessing though

Best