Difference between revisions of "CamCASP/Notes/7"

From CUC3
Jump to navigation Jump to search
import>Am592
 
import>Am592
Line 3: Line 3:
 
The ''${CAMCASP}/bin/'' folder contains a few batch scripts for running sets of jobs at various dimer geometries.
 
The ''${CAMCASP}/bin/'' folder contains a few batch scripts for running sets of jobs at various dimer geometries.
 
 
==SAPT(DFT) jobs with '''batch_SAPT.pl'''==
+
==SAPT(DFT) jobs with batch_SAPT.pl==
 
You cannot always perform an energy scan using the '''ENERGY-SCAN''' module in CamCASP as this module requires you to use the MC type of basis. If you need to use the MC+, DC or DC+ types of bases, then you will need to use the '''batch_SAPT.pl''' script. What this script does is to use a template CLUSTER command file and a file containing the translation and rotation parameters that define the dimer in order to run a batch of SAPT(DFT) (and also SAPT) calculations in an automatic way.
 
You cannot always perform an energy scan using the '''ENERGY-SCAN''' module in CamCASP as this module requires you to use the MC type of basis. If you need to use the MC+, DC or DC+ types of bases, then you will need to use the '''batch_SAPT.pl''' script. What this script does is to use a template CLUSTER command file and a file containing the translation and rotation parameters that define the dimer in order to run a batch of SAPT(DFT) (and also SAPT) calculations in an automatic way.
   
Line 13: Line 13:
 
-ipa IP(A) -ipb IP(B) [-q queue] [-M memory in MB]
 
-ipa IP(A) -ipb IP(B) [-q queue] [-M memory in MB]
 
</pre>
 
</pre>
  +
You need to supply:
  +
* The job name. For example, this might be ''benzene2''.
  +
* The name of the geometry file. More on this below. The default name will be ''jobname.geom'', so if this file is called ''benzene2.geom'', then there is no need to specify it explicitly.
  +
* The name of the CLUSTER command file. See below. As with the geometry file, there is a default value of ''jobname.clt''.
  +
* The type of basis to be used. (Is this now obsolete?)
  +
* The ionisation potentials of the two monomers.
  +
* The name of the queue. As always, there are special 'queues': ''bg'' says run the job in the background. The other special 'queue' ''none'' may not be usable.
  +
* You can also specify the memory to be used. This memory (in MB) will be passed to DALTON. The memory given to CamCASP should be in the CLUSTER file.

Revision as of 18:37, 17 March 2009

CamCASP => Notes => Using the batch scripts

The ${CAMCASP}/bin/ folder contains a few batch scripts for running sets of jobs at various dimer geometries.

SAPT(DFT) jobs with batch_SAPT.pl

You cannot always perform an energy scan using the ENERGY-SCAN module in CamCASP as this module requires you to use the MC type of basis. If you need to use the MC+, DC or DC+ types of bases, then you will need to use the batch_SAPT.pl script. What this script does is to use a template CLUSTER command file and a file containing the translation and rotation parameters that define the dimer in order to run a batch of SAPT(DFT) (and also SAPT) calculations in an automatic way.

Here are the options:

$ batch_SAPT.pl 
Usage:
    batch_SAPT.pl [-j] job [-geom name] [-clt name] [ -b dc | mc | mc+ | dc+ ]
        -ipa IP(A) -ipb IP(B) [-q queue] [-M memory in MB]

You need to supply:

  • The job name. For example, this might be benzene2.
  • The name of the geometry file. More on this below. The default name will be jobname.geom, so if this file is called benzene2.geom, then there is no need to specify it explicitly.
  • The name of the CLUSTER command file. See below. As with the geometry file, there is a default value of jobname.clt.
  • The type of basis to be used. (Is this now obsolete?)
  • The ionisation potentials of the two monomers.
  • The name of the queue. As always, there are special 'queues': bg says run the job in the background. The other special 'queue' none may not be usable.
  • You can also specify the memory to be used. This memory (in MB) will be passed to DALTON. The memory given to CamCASP should be in the CLUSTER file.