Pylada calls VASP as an external program: all it needs is to know how to call it. The path to the program can be configured in your ~/.pylada file by simply adding the following two lines:
>>> vasp_program = "/path/to/vasp"
>>> is_vasp_4 = True
vasp_program can be an absolute path, or simply the name of the VASP binary if it is available in you PATH environment variable. is_vasp_4 should be set to True or False depending on which version of VASP is available. It will prevent some vasp-5 only parameters from being set and will preferentially write the POSCAR in a vasp-5 format.
Note
It is also possible to tell the Vasp object to use a specific version:
>>> vasp = Vasp(program='/path/to/vasp')
It will apply only to calculations launched with that particular instance. If vasp does not have a program attribute, then it uses the global definition.
It is possible to run VASP in parallel. However, Pylada has to be set up correctly for running parallel code. Although it needs be done only once, it can be somewhat painful on some computers or supercomputers. Please see how to set up MPI calculations first.
The last configuration variable is verbose_representation. It controls whether or not the representation/print-out of the functional should include parameters which have not changed from the default. It is safest to keep it True.
See also