feat(dftb): add output directory support for SK table generation

Add output directory parameter to DFT2SKTable constructor and implement file path handling. This allows specifying where generated SK tables should be saved, creating the directory if it doesn't exist.

Author: QG-phy

dptb/nn/dftb/atomicdft.py

@@ -13,11 +13,12 @@
 from dptb.nn.sktb.builtin_skbasisDB import skbasisDB
 from dptb.nn.sktb.builtin_skbasisDB import onsite_e_builtin_basis, occupations_builtin_basis, Hubbard_U_builtin_basis
 import logging
+import os
 log = logging.getLogger(__name__)
 
 
 class DFT2SKTable(object):
-    def __init__(self, xc: str, superposition: str = 'density', scalarrel=True, **kwargs):
+    def __init__(self, xc: str, superposition: str = 'density', scalarrel:bool=True, output:str='./', **kwargs):
         '''This Python function initializes parameters for XC functional calculations, allowing for
         customization of the functional, superposition scheme, and scalar relativistic corrections.
         
@@ -43,7 +44,10 @@ def __init__(self, xc: str, superposition: str = 'density', scalarrel=True, **kw
         self.scalarrel = scalarrel
         self.kwargs = kwargs
         self.atomic_objs = {}
-            
+        if not os.path.exists(output):
+            os.makedirs(output)
+        self.output = output
+        self.out_filename_template = output + '/{el1}-{el2}.skf'
     def update_config(self, basis:Dict[str, List], rw: dict, pw:dict,  rd:dict=None, pd:dict=None):
         atomic_symbols = list(basis.keys())
         # check basis
@@ -133,15 +137,16 @@ def get_skf_pair(self, atom_a:str, atom_b:str=None, rmin=0.4, dr=0.02, N = 800,
         # Write the SK tables without repulsion (only electronic part)
         # for homo-case, the atomic eigenvalues, Hubbardvalues, occupations are also saved 
         # but the spe as well as the (spin-polarization error) is set to 0.0.
-    
+        
         if atom_a == atom_b:    
             self.off2c.write(eigenvalues=self.atomic_objs[atom_a]['eigenvalues'],
                      hubbardvalues=self.atomic_objs[atom_a]['hubbardvalues'],
                      occupations=self.atomic_objs[atom_a]['occupations'], 
-                     spe=0.
+                     spe=0.,
+                     filename_template=self.out_filename_template
                     )
         else:
-            self.off2c.write()
+            self.off2c.write(filename_template=self.out_filename_template)
         log.info(f'finished writing sk tables....')
 
     def get_full_pair(self, atom_a:str, atom_b:str=None, rmin=0.4, dr=0.02, N = 800, stride=1):