valyte 0.1.7__py3-none-any.whl → 0.1.9__py3-none-any.whl

valyte/band.py

@@ -3,19 +3,35 @@ Band structure KPOINTS generation module for Valyte.
 """
 
 import os
+import json
+import numpy as np
+import seekpath
+import spglib
 from pymatgen.core import Structure
 from pymatgen.symmetry.bandstructure import HighSymmKpath
+from pymatgen.symmetry.analyzer import SpacegroupAnalyzer
+try:
+    from importlib.resources import files as ilr_files
+except ImportError:
+    import importlib_resources as ilr_files
 
+from valyte.potcar import generate_potcar
 
-def generate_band_kpoints(poscar_path="POSCAR", npoints=40, output="KPOINTS"):
+
+def generate_band_kpoints(poscar_path="POSCAR", npoints=40, output="KPOINTS", symprec=0.01, mode="bradcrack"):
     """
     Generates KPOINTS file in line-mode for band structure calculations.
     Uses SeeK-path method for high-symmetry path determination.
     
+    IMPORTANT: Writes a standardized POSCAR (POSCAR_standard) that MUST be used
+    for the band structure calculation to ensure K-points are valid.
+    
     Args:
         poscar_path (str): Path to input POSCAR file.
         npoints (int): Number of points per segment (default: 40).
         output (str): Output filename for KPOINTS.
+        symprec (float): Symmetry precision for standardization (default: 0.01).
+        mode (str): Standardization convention (default: "bradcrack").
     """
     
     if not os.path.exists(poscar_path):
@@ -24,33 +40,250 @@ def generate_band_kpoints(poscar_path="POSCAR", npoints=40, output="KPOINTS"):
     # Read structure
     structure = Structure.from_file(poscar_path)
     
-    # Get high-symmetry path using SeeK-path method
-    kpath = HighSymmKpath(structure, path_type="setyawan_curtarolo")
-    
-    # Get the path
-    path = kpath.kpath["path"]
-    kpoints = kpath.kpath["kpoints"]
-    
+    # --- K-Point Generation Logic ---
+    if mode == "bradcrack":
+        try:
+            kpath = BradCrackKpath(structure, symprec=symprec)
+            prim_std = kpath.prim
+            path = kpath.path
+            kpoints = kpath.kpoints
+            
+            # Write standardized POSCAR from BradCrack logic
+            standard_filename = "POSCAR_standard"
+            prim_std.to(filename=standard_filename)
+        except Exception as e:
+            print(f"❌ Error generating BradCrack path: {e}")
+            return
+
+    else: 
+        # Fallback to Pymatgen logic for other modes
+        try:
+            # Map 'seekpath' alias to 'hinuma' which pymatgen uses (wrapper around seekpath)
+            if mode == "seekpath":
+                mode = "hinuma"
+
+            # Standardize structure first using SpacegroupAnalyzer
+            sga = SpacegroupAnalyzer(structure, symprec=symprec)
+            prim_std = sga.get_primitive_standard_structure()
+        except Exception as e:
+            print(f"❌ Error during standardization: {e}")
+            return
+
+        # Get high-symmetry path for the STANDARDIZED structure
+        try:
+            kpath = HighSymmKpath(prim_std, path_type=mode, symprec=symprec)
+            
+            # Write the standardized primitive structure
+            standard_filename = "POSCAR_standard"
+            prim_std.to(filename=standard_filename)
+            
+            # Get the path
+            path = kpath.kpath["path"]
+            kpoints = kpath.kpath["kpoints"]
+        except Exception as e:
+            print(f"❌ Error generating K-path: {e}")
+            return
+
     # Write KPOINTS file
-    with open(output, 'w') as f:
-        f.write("k-points for band structure\n")
-        f.write(f"{npoints}\n")
-        f.write("Line-mode\n")
-        f.write("Reciprocal\n")
-        
-        # Write each segment
-        for segment in path:
-            for i in range(len(segment) - 1):
-                start = segment[i]
-                end = segment[i + 1]
-                
-                start_coords = kpoints[start]
-                end_coords = kpoints[end]
-                
-                f.write(f"  {start_coords[0]:.6f}  {start_coords[1]:.6f}  {start_coords[2]:.6f}  ! {start}\n")
-                f.write(f"  {end_coords[0]:.6f}  {end_coords[1]:.6f}  {end_coords[2]:.6f}  ! {end}\n")
-                f.write("\n")
-    
-    # Print success message
-    path_str = ' → '.join([' - '.join(seg) for seg in path])
-    print(f"✅ KPOINTS generated: {output} ({path_str}, {npoints} pts/seg)")
+    try:
+        with open(output, "w") as f:
+            f.write("KPOINTS for Band Structure\n")
+            f.write(f"{npoints}\n")
+            f.write("Line-mode\n")
+            f.write("Reciprocal\n")
+            
+            for subpath in path:
+                for i in range(len(subpath) - 1):
+                    start_label = subpath[i]
+                    end_label = subpath[i+1]
+                    
+                    start_coords = kpoints[start_label]
+                    end_coords = kpoints[end_label]
+                    
+                    f.write(f"{start_coords[0]:10.6f} {start_coords[1]:10.6f} {start_coords[2]:10.6f} ! {start_label}\n")
+                    f.write(f"{end_coords[0]:10.6f} {end_coords[1]:10.6f} {end_coords[2]:10.6f} ! {end_label}\n")
+                    f.write("\n") # Optional newline between segments
+
+        print(f"✅ Generated {output} ({' - '.join([' - '.join(seg) for seg in path])})")
+        print(f"✅ Generated {standard_filename} (Standardized Primitive Cell)")
+        print(f"\n⚠️  IMPORTANT: You MUST use '{standard_filename}' for your band calculation!")
+        print(f"   The K-points are generated for this standardized orientation.")
+        print(f"   Using your original POSCAR may result in incorrect paths or 'Reciprocal lattice' errors.")
+
+    except Exception as e:
+        print(f"❌ Error writing KPOINTS file: {e}")
+
+    # --- POTCAR Generation ---
+    try:
+        print("ℹ️  Generating default POTCAR (PBE)...")
+        generate_potcar(poscar_path=poscar_path, functional="PBE", output="POTCAR")
+    except Exception as e:
+        print(f"⚠️  Could not generate POTCAR: {e}")
+        print("   (Proceeding without stopping, as KPOINTS are already generated)")
+
+
+
+class BradCrackKpath:
+    """
+    Native implementation of Bradley-Cracknell K-path generation.
+    Replicates logic from Sumo/SeeK-path to determine standard paths.
+    """
+    def __init__(self, structure, symprec=0.01):
+        self.structure = structure
+        self.symprec = symprec
+        
+        # Use SpacegroupAnalyzer for basic data
+        sga = SpacegroupAnalyzer(structure, symprec=symprec)
+        self._spg_data = sga.get_symmetry_dataset()
+        
+        # Use SeeK-path to get primitive/conventional structures matches Sumo Kpath.__init__
+        
+        # refine_cell logic from Sumo base class
+        # atom_numbers = [site.specie.number for site in structure] 
+        # But pymatgen structure to spglib cell tuple:
+        # cell = (lattice, positions, numbers)
+        cell = (structure.lattice.matrix, structure.frac_coords, [s.specie.number for s in structure])
+        
+        # Sumo uses spglib.refine_cell on the cell first? 
+        # "std = spglib.refine_cell(sym._cell, symprec=symprec)" 
+        # pymatgen sga._cell is (lattice, positions, numbers)
+        
+        # seekpath.get_path takes the cell structure
+        # output is dictionary
+        self._seek_data = seekpath.get_path(cell, symprec=symprec)
+        
+        # Reconstruct primitive structure from seekpath output
+        prim_lattice = self._seek_data["primitive_lattice"]
+        prim_pos = self._seek_data["primitive_positions"]
+        prim_types = self._seek_data["primitive_types"]
+        # Map types back to species? 
+        # We need a map from number to Element. 
+        # unique_species from sga?
+        # Let's just use explicit element list from input structure, assuming types are consistent?
+        # Or better, use sga to map Z to elements.
+        
+        # Setup element mapping
+        # Create a map from atomic number to Element object from input structure
+        z_to_specie = {s.specie.number: s.specie for s in structure}
+        prim_species = [z_to_specie[z] for z in prim_types]
+        
+        self.prim = Structure(prim_lattice, prim_species, prim_pos)
+        
+        conv_lattice = self._seek_data["conv_lattice"]
+        conv_pos = self._seek_data["conv_positions"]
+        conv_types = self._seek_data["conv_types"]
+        conv_species = [z_to_specie[z] for z in conv_types]
+        self.conv = Structure(conv_lattice, conv_species, conv_pos)
+        
+        # Now determine Bravais lattice for BradCrack
+        self._get_bradcrack_path()
+
+    def _get_bradcrack_path(self):
+        
+        # Determine lattice parameters from CONVENTIONAL cell
+        a, b, c = self.conv.lattice.abc
+        angles = self.conv.lattice.angles
+        # finding unique axis for monoclinic
+        # logic from BradCrackKpath.__init__
+        # "unique = angles.index(min(angles, key=angles.count))"
+        # usually 90, 90, beta. So unique is beta (non-90) index? No.
+        # Monoclinic: alpha=gamma=90, beta!=90. 90 appears twice. non-90 appears once.
+        # min count of angle values?
+        # if angles are [90, 90, 105], counts are {90:2, 105:1}. min count is 1. value is 105. index is 2.
+        # so unique is index of non-90 degree angle.
+        
+        # Round angles to avoid float issues
+        angles_r = [round(x, 3) for x in angles]
+        unique_val = min(angles_r, key=angles_r.count)
+        unique = angles_r.index(unique_val)
+
+        # Get Space Group Symbol and Number
+        # From seekpath? or sga? 
+        # Sumo uses: "spg_symbol = self.spg_symbol" which is "self._spg_data['international']"
+        # spglib dataset returns 'international'
+        spg_symbol = self._spg_data["international"]
+        spg_number = self._spg_data["number"]
+        
+        lattice_type = self.get_lattice_type(spg_number)
+        
+        bravais = self._get_bravais_lattice(spg_symbol, lattice_type, a, b, c, unique)
+        
+        # Load JSON
+        
+        json_file = ilr_files("valyte.data").joinpath("bradcrack.json")
+        with open(json_file, 'r') as f:
+            data = json.load(f)
+            
+        if bravais not in data:
+            raise ValueError(f"Bravais lattice code '{bravais}' not found in BradCrack data.")
+            
+        self.bradcrack_data = data[bravais]
+        self.kpoints = self.bradcrack_data["kpoints"]
+        self.path = self.bradcrack_data["path"]
+
+    def get_lattice_type(self, number):
+        # Logic from Sumo
+        if 1 <= number <= 2: return "triclinic"
+        if 3 <= number <= 15: return "monoclinic"
+        if 16 <= number <= 74: return "orthorhombic"
+        if 75 <= number <= 142: return "tetragonal"
+        if 143 <= number <= 167: 
+            if number in [146, 148, 155, 160, 161, 166, 167]: return "rhombohedral"
+            return "trigonal"
+        if 168 <= number <= 194: return "hexagonal"
+        if 195 <= number <= 230: return "cubic"
+        return "unknown"
+
+    def _get_bravais_lattice(self, spg_symbol, lattice_type, a, b, c, unique):
+        # Logic from Sumo BradCrackKpath._get_bravais_lattice
+        if lattice_type == "triclinic": return "triclinic"
+        
+        elif lattice_type == "monoclinic":
+            if "P" in spg_symbol:
+                if unique == 0: return "mon_p_a"
+                elif unique == 1: return "mon_p_b"
+                elif unique == 2: return "mon_p_c"
+            elif "C" in spg_symbol:
+                if unique == 0: return "mon_c_a"
+                elif unique == 1: return "mon_c_b"
+                elif unique == 2: return "mon_c_c"
+        
+        elif lattice_type == "orthorhombic":
+            if "P" in spg_symbol: return "orth_p"
+            elif "A" in spg_symbol or "C" in spg_symbol:
+                if a > b: return "orth_c_a"
+                elif b > a: return "orth_c_b"
+            elif "F" in spg_symbol:
+                # 1/a^2 etc conditions... need to replicate exact math
+                # Copied from Sumo source view
+                inv_a2 = 1/a**2; inv_b2 = 1/b**2; inv_c2 = 1/c**2
+                if (inv_a2 < inv_b2 + inv_c2) and (inv_b2 < inv_c2 + inv_a2) and (inv_c2 < inv_a2 + inv_b2):
+                    return "orth_f_1"
+                elif inv_c2 > inv_a2 + inv_b2: return "orth_f_2"
+                elif inv_b2 > inv_a2 + inv_c2: return "orth_f_3"
+                elif inv_a2 > inv_c2 + inv_b2: return "orth_f_4"
+            elif "I" in spg_symbol:
+                if a > b and a > c: return "orth_i_a"
+                elif b > a and b > c: return "orth_i_b"
+                elif c > a and c > b: return "orth_i_c"
+
+        elif lattice_type == "tetragonal":
+            if "P" in spg_symbol: return "tet_p"
+            elif "I" in spg_symbol:
+                if a > c: return "tet_i_a"
+                else: return "tet_i_c"
+
+        elif lattice_type in ["trigonal", "hexagonal", "rhombohedral"]:
+            if "R" in spg_symbol:
+                if a > np.sqrt(2)*c: return "trig_r_a"
+                else: return "trig_r_c"
+            elif "P" in spg_symbol:
+                if unique == 0: return "trig_p_a"
+                elif unique == 2: return "trig_p_c"
+
+        elif lattice_type == "cubic":
+            if "P" in spg_symbol: return "cubic_p"
+            elif "I" in spg_symbol: return "cubic_i"
+            elif "F" in spg_symbol: return "cubic_f"
+            
+        return "unknown"

valyte/cli.py

@@ -27,6 +27,7 @@ from valyte.band import generate_band_kpoints
 from valyte.band_plot import plot_band_structure
 from valyte.dos_plot import load_dos, plot_dos
 from valyte.kpoints import generate_kpoints_interactive
+from valyte.potcar import generate_potcar
 
 def parse_element_selection(inputs):
     """
@@ -116,10 +117,19 @@ def main():
     kpt_gen_parser.add_argument("-i", "--input", default="POSCAR", help="Input POSCAR file")
     kpt_gen_parser.add_argument("-n", "--npoints", type=int, default=40, help="Points per segment")
     kpt_gen_parser.add_argument("-o", "--output", default="KPOINTS", help="Output filename")
+    kpt_gen_parser.add_argument("--symprec", type=float, default=0.01, help="Symmetry precision (default: 0.01)")
+
+    kpt_gen_parser.add_argument("--mode", default="bradcrack", help="Standardization mode (default: bradcrack)")
 
     # --- K-Point Generation (Interactive) ---
     subparsers.add_parser("kpt", help="Interactive K-Point Generation (SCF)")
 
+    # --- POTCAR Generation ---
+    potcar_parser = subparsers.add_parser("potcar", help="Generate POTCAR")
+    potcar_parser.add_argument("-i", "--input", default="POSCAR", help="Input POSCAR file")
+    potcar_parser.add_argument("-o", "--output", default="POTCAR", help="Output filename")
+    potcar_parser.add_argument("--functional", default="PBE", help="Functional (default: PBE)")
+
     args = parser.parse_args()
 
     if args.command == "dos":
@@ -169,13 +179,26 @@ def main():
             print(f"❌ Error: {e}")
             sys.exit(1)
 
+    elif args.command == "potcar":
+        try:
+            generate_potcar(
+                poscar_path=args.input,
+                functional=args.functional,
+                output=args.output
+            )
+        except Exception as e:
+            print(f"❌ Error: {e}")
+            sys.exit(1)
+
     elif args.command == "band":
         if args.band_command == "kpt-gen":
             try:
                 generate_band_kpoints(
                     poscar_path=args.input,
                     npoints=args.npoints,
-                    output=args.output
+                    output=args.output,
+                    symprec=args.symprec,
+                    mode=args.mode
                 )
             except Exception as e:
                 print(f"❌ Error: {e}")