PDOS ratsph in version 7.8.1
Posted: Fri Jul 25, 2014 2:12 pm
Hi,
I just encountered a strange error when doing a calculation on our cluster, where we have abinit v 7.8.1 installed.
When running running a PDOS calculation with the following input parameters:
prtdos 3
natsph 1
iatsph 1
ratsph 3
I get the error message, that the "RATSPH" array should contain 4 values. This correlates with the number of atomic species in my calculation (4) rather than the number of atoms I want to calculate the PDOS for (1). Running the same code on my local computer with version 7.6.4 does not give this error.
After changing the input to "ratsph 3 2 2 2", the code runs through smoothly and prints the one PDOS I demanded.
Is this the (if I understand the helpfiles correctly) intended implementation of different radii for different atom sorts in PDOS and if yes, does this mean, I can set all the other values safely to any generic value? Furthermore, If I wanted to calculate the PDOS for an atom of the second atom type, would I then alter the second value rather than the first?
Thanks,
Joachim.
I just encountered a strange error when doing a calculation on our cluster, where we have abinit v 7.8.1 installed.
When running running a PDOS calculation with the following input parameters:
prtdos 3
natsph 1
iatsph 1
ratsph 3
I get the error message, that the "RATSPH" array should contain 4 values. This correlates with the number of atomic species in my calculation (4) rather than the number of atoms I want to calculate the PDOS for (1). Running the same code on my local computer with version 7.6.4 does not give this error.
After changing the input to "ratsph 3 2 2 2", the code runs through smoothly and prints the one PDOS I demanded.
Is this the (if I understand the helpfiles correctly) intended implementation of different radii for different atom sorts in PDOS and if yes, does this mean, I can set all the other values safely to any generic value? Furthermore, If I wanted to calculate the PDOS for an atom of the second atom type, would I then alter the second value rather than the first?
Thanks,
Joachim.