Implement Support for pymatgen.core.Molecule (#200)

* add .github/workflows/mirror_gitee.yml

* implement support for pymatgen.core.Molecule

* implement support for pymatgen.core.Molecule

* Change atom_types in dpdata/pymatgen/molecule.py

* Change doc string in dpdata/plugins/pymatgen.py

* add dpdata/pymatgin to setup.py

* modify molecule.py

* modify molecule.py

* modify molecule.py

* Support of pymatgen.Molecule in dpdata

* Update tests/test_pymatgen_molecule.py

* Moved "remove_pbc" from pymatgen/molecule.py to system.py

Author: tuoping

dpdata/plugins/pymatgen.py

@@ -1,4 +1,6 @@
 from dpdata.format import Format
+import dpdata.pymatgen.molecule
+import numpy as np
 
 
 @Format.register("pymatgen/structure")
@@ -22,6 +24,37 @@ def to_system(self, data, **kwargs):
         return structures
 
 
+@Format.register("pymatgen/molecule")
+class PyMatgenMoleculeFormat(Format):
+    @Format.post("remove_pbc")
+    def from_system(self, file_name, **kwargs):
+        try:
+            from pymatgen.core import Molecule
+        except ModuleNotFoundError as e:
+            raise ImportError('No module pymatgen.Molecule') from e
+
+        return dpdata.pymatgen.molecule.to_system_data(file_name)
+
+    def to_system(self, data, **kwargs):
+        """convert System to Pymatgen Molecule obj
+        """
+        molecules = []
+        try:
+            from pymatgen.core import Molecule
+        except ModuleNotFoundError as e:
+            raise ImportError('No module pymatgen.Molecule') from e
+
+        species = []
+        for name, numb in zip(data['atom_names'], data['atom_numbs']):
+            species.extend([name]*numb)
+        data = dpdata.system.remove_pbc(data)
+        for ii in range(np.array(data['coords']).shape[0]):
+            molecule = Molecule(
+                species, data['coords'][ii])
+            molecules.append(molecule)
+        return molecules
+
+
 @Format.register("pymatgen/computedstructureentry")
 @Format.register_to("to_pymatgen_ComputedStructureEntry")
 class PyMatgenCSEFormat(Format):
@@ -44,3 +77,5 @@ def to_labeled_system(self, data, *args, **kwargs):
             entry = ComputedStructureEntry(structure, energy, data=csedata)
             entries.append(entry)
         return entries
+
+        

dpdata/pymatgen/__init__.py

@@ -0,0 +1,26 @@
+import numpy as np
+from pymatgen.core import Molecule
+from collections import Counter
+import dpdata
+
+def to_system_data(file_name, protect_layer = 9) :
+    mol = Molecule.from_file(file_name)
+    elem_mol = list(str(site.species.elements[0]) for site in mol.sites)
+    elem_counter = Counter(elem_mol)
+    atom_names = list(elem_counter.keys())
+    atom_numbs = list(elem_counter.values())
+    atom_types = [list(atom_names).index(e) for e in elem_mol]
+    natoms = np.sum(atom_numbs)
+    
+    tmpcoord = np.copy(mol.cart_coords)
+
+    system = {}
+    system['atom_names'] = atom_names
+    system['atom_numbs'] = atom_numbs
+    system['atom_types'] = np.array(atom_types, dtype = int)
+    # center = [c - h_cell_size for c in mol.center_of_mass]
+    system['orig'] = np.array([0, 0, 0])
+
+    system['coords'] = [tmpcoord]
+    system['cells'] = [10.0 * np.eye(3)]
+    return system

dpdata/pymatgen/molecule.py

@@ -9,6 +9,7 @@
 from monty.serialization import loadfn,dumpfn
 from dpdata.periodic_table import Element
 from dpdata.amber.mask import pick_by_amber_mask, load_param_file
+import dpdata
 
 # ensure all plugins are loaded!
 import dpdata.plugins
@@ -418,7 +419,7 @@ def extend(self, systems):
         for system in systems:
             self.append(system.copy())
 
-
+    
     def apply_pbc(self) :
         """
         Append periodic boundary condition
@@ -428,6 +429,7 @@ def apply_pbc(self) :
         self.data['coords'] = np.matmul(ncoord, self.data['cells'])
 
 
+    @post_funcs.register("remove_pbc")
     def remove_pbc(self, protect_layer = 9):
         """
         This method does NOT delete the definition of the cells, it
@@ -441,19 +443,8 @@ def remove_pbc(self, protect_layer = 9):
         protect_layer : the protect layer between the atoms and the cell
                         boundary
         """
-        nframes = self.get_nframes()
-        natoms = self.get_natoms()
         assert(protect_layer >= 0), "the protect_layer should be no less than 0"
-        for ff in range(nframes):
-            tmpcoord = self.data['coords'][ff]
-            cog = np.average(tmpcoord, axis = 0)
-            dist = tmpcoord - np.tile(cog, [natoms, 1])
-            max_dist = np.max(np.linalg.norm(dist, axis = 1))
-            h_cell_size = max_dist + protect_layer
-            cell_size = h_cell_size * 2
-            shift = np.array([1,1,1]) * h_cell_size - cog
-            self.data['coords'][ff] = self.data['coords'][ff] + np.tile(shift, [natoms, 1])
-            self.data['cells'][ff] = cell_size * np.eye(3)
+        remove_pbc(self.data, protect_layer)
 
     def affine_map(self, trans, f_idx = 0) :
         assert(np.linalg.det(trans) != 0)
@@ -1314,3 +1305,18 @@ def elements_index_map(elements,standard=False,inverse=False):
     else:
         return dict(zip(elements,range(len(elements))))
 # %%
+
+def remove_pbc(system, protect_layer = 9):
+    nframes = len(system["coords"])
+    natoms = len(system['coords'][0])
+    for ff in range(nframes):
+        tmpcoord = system['coords'][ff]
+        cog = np.average(tmpcoord, axis = 0)
+        dist = tmpcoord - np.tile(cog, [natoms, 1])
+        max_dist = np.max(np.linalg.norm(dist, axis = 1))
+        h_cell_size = max_dist + protect_layer
+        cell_size = h_cell_size * 2
+        shift = np.array([1,1,1]) * h_cell_size - cog
+        system['coords'][ff] = system['coords'][ff] + np.tile(shift, [natoms, 1])
+        system['cells'][ff] = cell_size * np.eye(3)
+    return system

dpdata/system.py

@@ -41,6 +41,7 @@
               'dpdata/abacus',
               'dpdata/rdkit',
               'dpdata/plugins',
+              'dpdata/pymatgen',
     ],
     package_data={'dpdata':['*.json']},
     classifiers=[

setup.py

@@ -0,0 +1,16 @@
+C1 H5 N2 
+1.0
+2.2504659203492643e+01 0.0000000000000000e+00 0.0000000000000000e+00 
+0.0000000000000000e+00 2.2504659203492643e+01 0.0000000000000000e+00 
+0.0000000000000000e+00 0.0000000000000000e+00 2.2504659203492643e+01 
+C H N 
+1 5 2 
+Cartesian
+  11.2523296017   11.9178548890   11.2108422059
+  11.2523296017   11.4836546020   13.4873419539
+  11.2523296017    9.9657546120   12.5961418889
+  11.2523296017   11.5757547460    9.0233422059
+  11.2523296017    9.9209545690   10.0419421929
+  11.2523296017   13.0578542790   11.2108422059
+  11.2523296017   11.0671545110   12.4584418079
+  11.2523296017   11.0296546060    9.9897423529

tests/pymatgen/FA-001.vasp

@@ -0,0 +1,10 @@
+8
+Lattice="10   0.0   0.0   0.0   10   0.0   0.0   0.0   10"     Properties=species:S:1:pos:R:3:Z:I:1
+C    3.1742845125747904e-16  5.342900276  5.184000015         6
+H    4.568238577398768e-16  4.908699989  7.460499763          1
+N    3.938218942178196e-16  4.492199898  6.431599617          7
+H    4.022535923897687e-16  3.390799999  6.569299698          1
+H    1.834827076007373e-16  5.000800133  2.996500015          1
+N    2.4265764993669136e-16  4.454699993  3.962900162         7
+H    2.4585396828529157e-16  3.345999956  4.015100002         1
+H    3.1742845125747904e-16  6.482899666  5.184000015         1

tests/pymatgen/FA-001.xyz

@@ -0,0 +1,16 @@
+H5 C1 N2
+1.0
+10.000000 0.000000 0.000000
+0.000000 10.000000 0.000000
+0.000000 0.000000 10.000000
+H C N
+5 1 2
+direct
+0.577383 0.631337 0.500367 H
+0.440363 0.395639 0.623009 H
+0.523296 0.528867 0.297597 H
+0.521491 0.528803 0.702604 H
+0.441477 0.395690 0.376400 H
+0.521985 0.537275 0.500106 C
+0.490941 0.484747 0.383775 N
+0.489919 0.484705 0.616136 N

tests/pymatgen/mol2-new.vasp

@@ -0,0 +1,16 @@
+hexagonal-PbI3                          
+   1.00000000000000     
+     8.4261398403369476   -0.0207094818066008   -0.0031014314051307
+    -4.2326852079273163    7.4079517949230027    0.0021007606821090
+    -0.0033015270376727    0.0018001191064931    7.9750946862221301
+    N    C    H 
+        2     1    5
+Direct
+  0.3230984076298932  0.6454769144393403  0.1041651474646342
+  0.3220267174078323  0.6453279258574751  0.3955245499807875
+  0.3958479883967422  0.7164380537464232  0.2500334235810864
+  0.5257412318253072  0.8435919770693032  0.2503606409552951
+  0.2024793011832460  0.5249247262853158  0.4041424751787161
+  0.3915235842243226  0.7051542887367509  -0.0038936922276933 0.9961063077723067
+  0.3896311952726235  0.7049043576463694  0.5039461714481914
+  0.2036571496859804  0.5250922648687301  0.0949178563751044

tests/pymatgen/mol2.vasp

@@ -0,0 +1,31 @@
+import os
+import numpy as np
+import unittest
+from context import dpdata
+
+class TestPOSCARCart(unittest.TestCase):
+    
+    def setUp(self): 
+        self.system = dpdata.System()
+        self.system.from_pymatgen_molecule(os.path.join('pymatgen', 'FA-001.xyz'))
+        self.assertEqual(list(self.system["atom_types"]), [0, 1, 2, 1, 1, 2, 1, 1])
+
+    def test_poscar_to_molecule(self):
+        tmp_system = dpdata.System()
+        tmp_system.from_vasp_poscar(os.path.join('pymatgen', 'mol2.vasp'))
+        natoms = len(tmp_system['coords'][0])
+        tmpcoord = tmp_system['coords'][0]
+        cog = np.average(tmpcoord, axis = 0)
+        dist = tmpcoord - np.tile(cog, [natoms, 1])
+        max_dist_0 = np.max(np.linalg.norm(dist, axis = 1))
+
+        mols = tmp_system.to("pymatgen/molecule")
+        cog = np.average(mols[-1].cart_coords, axis = 0)
+        dist = mols[-1].cart_coords - np.tile(cog, [natoms, 1])
+        max_dist_1 = np.max(np.linalg.norm(dist, axis = 1))
+        self.assertAlmostEqual(max_dist_0, max_dist_1)
+        
+
+
+if __name__ == '__main__':
+    unittest.main()