ElPh Code Refactoring

[PuffyDucks]

Many changes being made for this PR.

Main Changes

• ElPh now sends Slurm jobs by default
• More print messages for currently running and skipped calculations
• --overwrite flag added
• Phasing out nmol hard coding

Details

Arguments

• Use -l to run ElPh on local machine
• Flags --gpu, --account, --time, --hpc have been added to modify Slum script generation
• --homo and --svd are commented out due to never being used in the code
• --overwrite flag added (explained in 'Workflow')
• main.py now validates --workflow and --nmol at start
  • workflow must be {1, 2, 3}. nmol is currently restricted to {3, 4} (can possibly go to 6 without bugs? untested.)

Workflow

• ElPh no longer skips full workflow when final files are present
• Individual steps are still skipped when their output files are detected
  • The --overwrite flag prevents any steps from being skipped
• Workflow skips and their file path are printed out for transparency
• Most calculations will print a message before and/or after running
• Above changes are currently implemented for workflow 1

Improving code quality

• Repetitive/hard-coded logic is to be replaced
  • The most common is if nmol == 3: ... if nmol == 4: ...
  • Functions are often called repeatedly with predictable changes to arguments
• For loops now used to make the code more flexible and easier to maintain
• Like above, changes are only committed to workflow 1 at the moment

pathlib.Path

• Workflow 1 code has been modified to use pathlib.Path in place of os.path
  • Not a huge deal, but makes the code nicer to read
  • A bit safer and easier

Error throwing

• ut.print_error(message) has been changed to ut.throw_error(message, exit_status=1)
  • the function now runs sys.exit(exit_status)

[Raykang35]

In mapping section, the atom order in POSCAR is different than the atom order in .cif. I am not 100% sure whether .cif file would return the same output as POSCAR. There are 2 ways we can deal with this.

1. I suggest here we only apply POSCAR. We should add a function that if .cif, convert into POSCAR.
2. It is worth testing with simple material simulation such as BTBT. If the output is the same, leave it as it is.

[Raykang35]

I don't think we need any geometry file from 2-phonons folder. The geometry file in 2-phonons is different than original geometry. It is optimized.

[Raykang35]

The ideas of improving the code quality and workflow is great. Let's improve workflow 2 and 3 as well.