chore: make more imports optional

examples/build_model_using_mofdscribe.ipynb

@@ -61,7 +61,6 @@
     "from pymatviz.parity import density_scatter_with_hist\n",
     "\n",
     "from mofdscribe.featurizers.chemistry import APRDF, RACS, AMD\n",
-    "from mofdscribe.featurizers.topology.ph_hist import PHHist\n",
     "from mofdscribe.featurizers.base import MOFMultipleFeaturizer\n",
     "from mofdscribe.datasets.thermal_stability_dataset import ThermalStabilityDataset\n",
     "from mofdscribe.datasets.structuredataset import FrameDataset\n",
@@ -192,7 +191,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "featurizer = MOFMultipleFeaturizer([APRDF(), PHHist(), RACS(), AMD()])"
+    "featurizer = MOFMultipleFeaturizer([APRDF(), RACS(), AMD()])"
    ]
   },
   {

src/mofdscribe/featurizers/bu/utils.py

@@ -394,7 +394,6 @@
 import numpy as np
 from loguru import logger
 from numpy.typing import ArrayLike
-
 from pymatgen.core import Molecule
 from rdkit import Chem
 from rdkit.Chem import BondType
@@ -452,7 +451,7 @@
 }
 
 
-def pymatgen_2_babel_atom_idx_map(pmg_mol: Molecule, ob_mol: "ob.OBMol") -> Dict[int, int]:
+def pymatgen_2_babel_atom_idx_map(pmg_mol: Molecule, ob_mol) -> Dict[int, int]:
     """
     Create an atom index mapping between pymatgen mol and openbabel mol.
 
@@ -471,6 +470,7 @@ def pymatgen_2_babel_atom_idx_map(pmg_mol: Molecule, ob_mol: "ob.OBMol") -> Dict
         RuntimeError: if mapping is not possible.
     """
     from openbabel import openbabel as ob
+
     pmg_coords = pmg_mol.cart_coords
     ob_coords = [[a.GetX(), a.GetY(), a.GetZ()] for a in ob.OBMolAtomIter(ob_mol)]
     ob_index = [a.GetIdx() for a in ob.OBMolAtomIter(ob_mol)]
@@ -644,6 +644,7 @@ def create_rdkit_mol_from_mol_graph(
     """
     from openbabel import openbabel as ob
     from pymatgen.io.babel import BabelMolAdaptor
+
     pymatgen_mol = mol_graph.molecule
     species = [str(s) for s in pymatgen_mol.species]
     coords = pymatgen_mol.cart_coords

src/mofdscribe/featurizers/topology/_tda_helpers.py

@@ -6,9 +6,6 @@
 import numpy as np
 from element_coder import encode_many
 from loguru import logger
-from moleculetda.construct_pd import construct_pds
-from moleculetda.read_file import make_supercell
-from moleculetda.vectorize_pds import diagrams_to_arrays, get_images
 from pymatgen.core import Structure
 from pymatgen.transformations.advanced_transformations import CubicSupercellTransformation
 
@@ -18,6 +15,8 @@
 
 # @np_cache
 def construct_pds_cached(coords, periodic=False, weights: Optional[Collection] = None):
+    from moleculetda.construct_pd import construct_pds
+
     return construct_pds(coords, periodic=periodic, weights=weights)
 
 
@@ -28,6 +27,7 @@ def _coords_for_structure(
     no_supercell: bool = False,
     weighting: Optional[str] = None,
 ) -> Tuple[np.ndarray, np.ndarray]:
+    from moleculetda.read_file import make_supercell
 
     if no_supercell:
         if weighting is not None:
@@ -64,6 +64,8 @@ def _coords_for_structure(
 def _pd_arrays_from_coords(
     coords, periodic: bool = False, bd_arrays: bool = False, weights: Optional[np.ndarray] = None
 ):
+    from moleculetda.vectorize_pds import diagrams_to_arrays
+
     pds = construct_pds_cached(coords, periodic=periodic, weights=weights)
     if bd_arrays:
         pd = diagrams_to_bd_arrays(pds)
@@ -118,6 +120,8 @@ def get_persistent_images_for_structure(
         persistent_images (dict): dictionary of persistent images and their
             barcode representations
     """
+    from moleculetda.vectorize_pds import get_images
+
     element_images: Dict[dict] = defaultdict(dict)
     specs = []
     for mb, mp in zip(max_b, max_p):

src/mofdscribe/featurizers/utils/eqeq/__init__.py

@@ -8,13 +8,14 @@
 from typing import List, Tuple
 
 from ase.io.cif import write_cif
-from pyeqeq import run_on_cif
 from pymatgen.core import IStructure
 from pymatgen.io.ase import AseAtomsAdaptor
 
 
 @lru_cache(maxsize=32)
 def get_eqeq_charges(structure: IStructure) -> Tuple[str, List[float]]:
+    from pyeqeq import run_on_cif
+
     with open(os.devnull, "w") as devnull, contextlib.redirect_stdout(devnull), NamedTemporaryFile(
         "w", suffix=".cif"
     ) as f: