structure_environments.py

"""
This module contains objects that are used to describe the environments in a structure. The most detailed object
(StructureEnvironments) contains a very thorough analysis of the environments of a given atom but is difficult to
used as such. The LightStructureEnvironments object is a lighter version that is obtained by applying a "strategy"
on the StructureEnvironments object. Basically, the LightStructureEnvironments provides the coordination environment(s)
and possibly some fraction corresponding to these.
"""

from __future__ import annotations

import matplotlib.pyplot as plt
import numpy as np
from matplotlib import cm
from matplotlib.colors import Normalize
from matplotlib.gridspec import GridSpec
from matplotlib.patches import Polygon
from monty.json import MontyDecoder, MSONable, jsanitize

from pymatgen.analysis.chemenv.coordination_environments.coordination_geometries import AllCoordinationGeometries
from pymatgen.analysis.chemenv.coordination_environments.voronoi import DetailedVoronoiContainer
from pymatgen.analysis.chemenv.utils.chemenv_errors import ChemenvError
from pymatgen.analysis.chemenv.utils.defs_utils import AdditionalConditions
from pymatgen.core import Element, PeriodicNeighbor, PeriodicSite, Species, Structure

__author__ = "David Waroquiers"
__copyright__ = "Copyright 2012, The Materials Project"
__credits__ = "Geoffroy Hautier"
__version__ = "2.0"
__maintainer__ = "David Waroquiers"
__email__ = "david.waroquiers@gmail.com"
__date__ = "Feb 20, 2016"


all_cg = AllCoordinationGeometries()
symbol_cn_mapping = all_cg.get_symbol_cn_mapping()


class StructureEnvironments(MSONable):
    """Class used to store the chemical environments of a given structure."""

    AC = AdditionalConditions()

    class NeighborsSet:
        """Class used to store a given set of neighbors of a given site (based on the detailed_voronoi)."""

        def __init__(self, structure: Structure, isite, detailed_voronoi, site_voronoi_indices, sources=None):
            """Constructor for NeighborsSet.

            Args:
                structure: Structure object.
                isite: Index of the site for which neighbors are stored in this NeighborsSet.
                detailed_voronoi: Corresponding DetailedVoronoiContainer object containing all the possible
                    neighbors of the give site.
                site_voronoi_indices: Indices of the voronoi sites in the DetailedVoronoiContainer object that
                    make up this NeighborsSet.
                sources: Sources for this NeighborsSet, i.e. how this NeighborsSet was generated.
            """
            self.structure = structure
            self.isite = isite
            self.detailed_voronoi = detailed_voronoi
            self.voronoi = detailed_voronoi.voronoi_list2[isite]
            if n_dupes := (len(set(site_voronoi_indices)) - len(site_voronoi_indices)):
                raise ValueError(f"Set of neighbors contains {n_dupes} duplicates!")
            self.site_voronoi_indices = sorted(site_voronoi_indices)
            if sources is None:
                self.sources = [{"origin": "UNKNOWN"}]
            elif isinstance(sources, list):
                self.sources = sources
            else:
                self.sources = [sources]

        def get_neighb_voronoi_indices(self, permutation):
            """Get indices in the detailed_voronoi corresponding to the current permutation.

            Args:
                permutation: Current permutation for which the indices in the detailed_voronoi are needed.

            Returns:
                list[int]: indices in the detailed_voronoi.
            """
            return [self.site_voronoi_indices[ii] for ii in permutation]

        @property
        def neighb_coords(self):
            """Coordinates of neighbors for this NeighborsSet."""
            return [self.voronoi[inb]["site"].coords for inb in self.site_voronoi_indices]

        @property
        def neighb_coordsOpt(self):
            """Optimized access to the coordinates of neighbors for this NeighborsSet."""
            return self.detailed_voronoi.voronoi_list_coords[self.isite].take(self.site_voronoi_indices, axis=0)

        @property
        def neighb_sites(self):
            """Neighbors for this NeighborsSet as pymatgen Sites."""
            return [self.voronoi[inb]["site"] for inb in self.site_voronoi_indices]

        @property
        def neighb_sites_and_indices(self):
            """List of neighbors for this NeighborsSet as pymatgen Sites and their index in the original structure."""
            return [
                {"site": self.voronoi[inb]["site"], "index": self.voronoi[inb]["index"]}
                for inb in self.site_voronoi_indices
            ]

        @property
        def coords(self):
            """Coordinates of the current central atom and its neighbors for this NeighborsSet."""
            coords = [self.structure[self.isite].coords]
            coords.extend(self.neighb_coords)
            return coords

        @property
        def normalized_distances(self):
            """Normalized distances to each neighbor in this NeighborsSet."""
            return [self.voronoi[inb]["normalized_distance"] for inb in self.site_voronoi_indices]

        @property
        def normalized_angles(self):
            """Normalized angles for each neighbor in this NeighborsSet."""
            return [self.voronoi[inb]["normalized_angle"] for inb in self.site_voronoi_indices]

        @property
        def distances(self):
            """Distances to each neighbor in this NeighborsSet."""
            return [self.voronoi[inb]["distance"] for inb in self.site_voronoi_indices]

        @property
        def angles(self):
            """Angles for each neighbor in this NeighborsSet."""
            return [self.voronoi[inb]["angle"] for inb in self.site_voronoi_indices]

        # @property
        # def sphere_fraction_angles(self):
        #     return [0.25 * self.voronoi[inb]['angle'] / np.pi for inb in self.site_voronoi_indices]

        @property
        def info(self):
            """Summarized information about this NeighborsSet."""
            was = self.normalized_angles
            wds = self.normalized_distances
            angles = self.angles
            distances = self.distances
            return {
                "normalized_angles": was,
                "normalized_distances": wds,
                "normalized_angles_sum": np.sum(was),
                "normalized_angles_mean": np.mean(was),
                "normalized_angles_std": np.std(was),
                "normalized_angles_min": np.min(was),
                "normalized_angles_max": np.max(was),
                "normalized_distances_mean": np.mean(wds),
                "normalized_distances_std": np.std(wds),
                "normalized_distances_min": np.min(wds),
                "normalized_distances_max": np.max(wds),
                "angles": angles,
                "distances": distances,
                "angles_sum": np.sum(angles),
                "angles_mean": np.mean(angles),
                "angles_std": np.std(angles),
                "angles_min": np.min(angles),
                "angles_max": np.max(angles),
                "distances_mean": np.mean(distances),
                "distances_std": np.std(distances),
                "distances_min": np.min(distances),
                "distances_max": np.max(distances),
            }

        def distance_plateau(self):
            """Returns the distances plateau's for this NeighborsSet."""
            all_nbs_normalized_distances_sorted = sorted(
                (nb["normalized_distance"] for nb in self.voronoi), reverse=True
            )
            maxdist = np.max(self.normalized_distances)
            plateau = None
            for idist, dist in enumerate(all_nbs_normalized_distances_sorted):
                if np.isclose(
                    dist,
                    maxdist,
                    rtol=0.0,
                    atol=self.detailed_voronoi.normalized_distance_tolerance,
                ):
                    plateau = np.inf if idist == 0 else all_nbs_normalized_distances_sorted[idist - 1] - maxdist
                    break
            if plateau is None:
                raise ValueError("Plateau not found ...")
            return plateau

        def angle_plateau(self):
            """Returns the angles plateau's for this NeighborsSet."""
            all_nbs_normalized_angles_sorted = sorted(nb["normalized_angle"] for nb in self.voronoi)
            minang = np.min(self.normalized_angles)
            for nb in self.voronoi:
                print(nb)
            plateau = None
            for iang, ang in enumerate(all_nbs_normalized_angles_sorted):
                if np.isclose(
                    ang,
                    minang,
                    rtol=0.0,
                    atol=self.detailed_voronoi.normalized_angle_tolerance,
                ):
                    plateau = minang if iang == 0 else minang - all_nbs_normalized_angles_sorted[iang - 1]
                    break
            if plateau is None:
                raise ValueError("Plateau not found ...")
            return plateau

        def voronoi_grid_surface_points(self, additional_condition=1, other_origins="DO_NOTHING"):
            """
            Get the surface points in the Voronoi grid for this neighbor from the sources.
            The general shape of the points should look like a staircase such as in the following figure :

               ^
            0.0|
               |
               |      B----C
               |      |    |
               |      |    |
            a  |      k    D-------E
            n  |      |            |
            g  |      |            |
            l  |      |            |
            e  |      j            F----n---------G
               |      |                           |
               |      |                           |
               |      A----g-------h----i---------H
               |
               |
            1.0+------------------------------------------------->
              1.0              distance              2.0   ->+Inf

            Args:
                additional_condition: Additional condition for the neighbors.
                other_origins: What to do with sources that do not come from the Voronoi grid (e.g. "from hints").
            """
            src_list = []
            for src in self.sources:
                if src["origin"] == "dist_ang_ac_voronoi":
                    if src["ac"] != additional_condition:
                        continue
                    src_list.append(src)
                else:
                    if other_origins == "DO_NOTHING":
                        continue
                    raise NotImplementedError("Nothing implemented for other sources ...")
            if len(src_list) == 0:
                return None

            dists = [src["dp_dict"]["min"] for src in src_list]
            angles = [src["ap_dict"]["max"] for src in src_list]
            next_dists = [src["dp_dict"]["next"] for src in src_list]
            next_angles = [src["ap_dict"]["next"] for src in src_list]

            points_dict = {}

            p_dists = []
            pangs = []

            for idx in range(len(src_list)):
                if not any(np.isclose(p_dists, dists[idx])):
                    p_dists.append(dists[idx])
                if not any(np.isclose(p_dists, next_dists[idx])):
                    p_dists.append(next_dists[idx])
                if not any(np.isclose(pangs, angles[idx])):
                    pangs.append(angles[idx])
                if not any(np.isclose(pangs, next_angles[idx])):
                    pangs.append(next_angles[idx])
                d1_indices = np.argwhere(np.isclose(p_dists, dists[idx])).flatten()
                if len(d1_indices) != 1:
                    raise ValueError("Distance parameter not found ...")
                d2_indices = np.argwhere(np.isclose(p_dists, next_dists[idx])).flatten()
                if len(d2_indices) != 1:
                    raise ValueError("Distance parameter not found ...")
                a1_indices = np.argwhere(np.isclose(pangs, angles[idx])).flatten()
                if len(a1_indices) != 1:
                    raise ValueError("Angle parameter not found ...")
                a2_indices = np.argwhere(np.isclose(pangs, next_angles[idx])).flatten()
                if len(a2_indices) != 1:
                    raise ValueError("Angle parameter not found ...")
                id1 = d1_indices[0]
                id2 = d2_indices[0]
                ia1 = a1_indices[0]
                ia2 = a2_indices[0]
                for id_ia in [(id1, ia1), (id1, ia2), (id2, ia1), (id2, ia2)]:
                    points_dict.setdefault(id_ia, 0)
                    points_dict[id_ia] += 1

            new_pts = []
            for pt, pt_nb in points_dict.items():
                if pt_nb % 2 == 1:
                    new_pts.append(pt)

            sorted_points = [(0, 0)]
            move_ap_index = True
            while True:
                last_pt = sorted_points[-1]
                if move_ap_index:  # "Move" the angle parameter
                    idp = last_pt[0]
                    iap = None
                    for pt in new_pts:
                        if pt[0] == idp and pt != last_pt:
                            iap = pt[1]
                            break
                else:  # "Move" the distance parameter
                    idp = None
                    iap = last_pt[1]
                    for pt in new_pts:
                        if pt[1] == iap and pt != last_pt:
                            idp = pt[0]
                            break
                if (idp, iap) == (0, 0):
                    break
                if (idp, iap) in sorted_points:
                    raise ValueError("Error sorting points ...")
                sorted_points.append((idp, iap))
                move_ap_index = not move_ap_index

            return [(p_dists[idp], pangs[iap]) for (idp, iap) in sorted_points]

        @property
        def source(self):
            """
            Returns the source of this NeighborsSet (how it was generated, e.g. from which Voronoi
            cutoffs, or from hints).
            """
            if len(self.sources) != 1:
                raise RuntimeError("Number of sources different from 1 !")
            return self.sources[0]

        def add_source(self, source):
            """
            Add a source to this NeighborsSet.

            Args:
                source: Information about the generation of this NeighborsSet.
            """
            if source not in self.sources:
                self.sources.append(source)

        def __len__(self):
            return len(self.site_voronoi_indices)

        def __hash__(self) -> int:
            return len(self.site_voronoi_indices)

        def __eq__(self, other: object) -> bool:
            needed_attrs = ("isite", "site_voronoi_indices")
            if not all(hasattr(other, attr) for attr in needed_attrs):
                return NotImplemented
            return all(getattr(self, attr) == getattr(other, attr) for attr in needed_attrs)

        def __str__(self):
            out = f"Neighbors Set for site #{self.isite} :\n"
            out += f" - Coordination number : {len(self)}\n"
            voro_indices = ", ".join(f"{site_voronoi_index}" for site_voronoi_index in self.site_voronoi_indices)
            out += f" - Voronoi indices : {voro_indices}\n"
            return out

        def as_dict(self):
            """A JSON-serializable dict representation of the NeighborsSet."""
            return {
                "isite": self.isite,
                "site_voronoi_indices": self.site_voronoi_indices,
                "sources": self.sources,
            }

        @classmethod
        def from_dict(cls, dct, structure: Structure, detailed_voronoi):
            """
            Reconstructs the NeighborsSet algorithm from its JSON-serializable dict representation, together with
            the structure and the DetailedVoronoiContainer.

            As an inner (nested) class, the NeighborsSet is not supposed to be used anywhere else that inside the
            StructureEnvironments. The from_dict method is thus using the structure and  detailed_voronoi when
            reconstructing itself. These two are both in the StructureEnvironments object.

            Args:
                dct: a JSON-serializable dict representation of a NeighborsSet.
                structure: The structure.
                detailed_voronoi: The Voronoi object containing all the neighboring atoms from which the subset of
                    neighbors for this NeighborsSet is extracted.

            Returns:
                NeighborsSet
            """
            return cls(
                structure=structure,
                isite=dct["isite"],
                detailed_voronoi=detailed_voronoi,
                site_voronoi_indices=dct["site_voronoi_indices"],
                sources=dct["sources"],
            )

    def __init__(
        self,
        voronoi,
        valences,
        sites_map,
        equivalent_sites,
        ce_list,
        structure,
        neighbors_sets=None,
        info=None,
    ):
        """
        Constructor for the StructureEnvironments object.

        Args:
            voronoi: VoronoiContainer object for the structure.
            valences: Valences provided.
            sites_map: Mapping of equivalent sites to the nonequivalent sites that have been computed.
            equivalent_sites: List of list of equivalent sites of the structure.
            ce_list: List of chemical environments.
            structure: Structure object.
            neighbors_sets: List of neighbors sets.
            info: Additional information for this StructureEnvironments object.
        """
        self.voronoi = voronoi
        self.valences = valences
        self.sites_map = sites_map
        self.equivalent_sites = equivalent_sites
        # self.struct_sites_to_irreducible_site_list_map = struct_sites_to_irreducible_site_list_map
        self.ce_list = ce_list
        self.structure = structure
        if neighbors_sets is None:
            self.neighbors_sets = [None] * len(self.structure)
        else:
            self.neighbors_sets = neighbors_sets
        self.info = info

    def init_neighbors_sets(self, isite, additional_conditions=None, valences=None):
        """
        Initialize the list of neighbors sets for the current site.

        Args:
            isite: Index of the site under consideration.
            additional_conditions: Additional conditions to be used for the initialization of the list of
                neighbors sets, e.g. "Only anion-cation bonds", ...
            valences: List of valences for each site in the structure (needed if an additional condition based on the
                valence is used, e.g. only anion-cation bonds).
        """
        site_voronoi = self.voronoi.voronoi_list2[isite]
        if site_voronoi is None:
            return
        if additional_conditions is None:
            additional_conditions = self.AC.ALL
        if (self.AC.ONLY_ACB in additional_conditions or self.AC.ONLY_ACB_AND_NO_E2SEB) and valences is None:
            raise ChemenvError(
                "StructureEnvironments",
                "init_neighbors_sets",
                "Valences are not given while only_anion_cation_bonds are allowed. Cannot continue",
            )
        site_distance_parameters = self.voronoi.neighbors_normalized_distances[isite]
        site_angle_parameters = self.voronoi.neighbors_normalized_angles[isite]
        # Precompute distance conditions
        distance_conditions = []
        for idp, dp_dict in enumerate(site_distance_parameters):
            distance_conditions.append([])
            for inb in range(len(site_voronoi)):
                cond = inb in dp_dict["nb_indices"]
                distance_conditions[idp].append(cond)
        # Precompute angle conditions
        angle_conditions = []
        for iap, ap_dict in enumerate(site_angle_parameters):
            angle_conditions.append([])
            for inb in range(len(site_voronoi)):
                cond = inb in ap_dict["nb_indices"]
                angle_conditions[iap].append(cond)
        # Precompute additional conditions
        precomputed_additional_conditions = {ac: [] for ac in additional_conditions}
        for voro_nb_dict in site_voronoi:
            for ac in additional_conditions:
                cond = self.AC.check_condition(
                    condition=ac,
                    structure=self.structure,
                    parameters={
                        "valences": valences,
                        "neighbor_index": voro_nb_dict["index"],
                        "site_index": isite,
                    },
                )
                precomputed_additional_conditions[ac].append(cond)
        # Add the neighbors sets based on the distance/angle/additional parameters
        for idp, dp_dict in enumerate(site_distance_parameters):
            for iap, ap_dict in enumerate(site_angle_parameters):
                for iac, ac in enumerate(additional_conditions):
                    src = {
                        "origin": "dist_ang_ac_voronoi",
                        "idp": idp,
                        "iap": iap,
                        "dp_dict": dp_dict,
                        "ap_dict": ap_dict,
                        "iac": iac,
                        "ac": ac,
                        "ac_name": self.AC.CONDITION_DESCRIPTION[ac],
                    }
                    site_voronoi_indices = [
                        inb
                        for inb, voro_nb_dict in enumerate(site_voronoi)
                        if (
                            distance_conditions[idp][inb]
                            and angle_conditions[iap][inb]
                            and precomputed_additional_conditions[ac][inb]
                        )
                    ]
                    nb_set = self.NeighborsSet(
                        structure=self.structure,
                        isite=isite,
                        detailed_voronoi=self.voronoi,
                        site_voronoi_indices=site_voronoi_indices,
                        sources=src,
                    )
                    self.add_neighbors_set(isite=isite, nb_set=nb_set)

    def add_neighbors_set(self, isite, nb_set):
        """
        Adds a neighbor set to the list of neighbors sets for this site.

        Args:
            isite: Index of the site under consideration.
            nb_set: NeighborsSet to be added.
        """
        if self.neighbors_sets[isite] is None:
            self.neighbors_sets[isite] = {}
            self.ce_list[isite] = {}
        cn = len(nb_set)
        if cn not in self.neighbors_sets[isite]:
            self.neighbors_sets[isite][cn] = []
            self.ce_list[isite][cn] = []
        try:
            nb_set_index = self.neighbors_sets[isite][cn].index(nb_set)
            self.neighbors_sets[isite][cn][nb_set_index].add_source(nb_set.source)
        except ValueError:
            self.neighbors_sets[isite][cn].append(nb_set)
            self.ce_list[isite][cn].append(None)

    def update_coordination_environments(self, isite, cn, nb_set, ce):
        """
        Updates the coordination environment for this site, coordination and neighbor set.

        Args:
            isite: Index of the site to be updated.
            cn: Coordination to be updated.
            nb_set: Neighbors set to be updated.
            ce: ChemicalEnvironments object for this neighbors set.
        """
        if self.ce_list[isite] is None:
            self.ce_list[isite] = {}
        if cn not in self.ce_list[isite]:
            self.ce_list[isite][cn] = []
        try:
            nb_set_index = self.neighbors_sets[isite][cn].index(nb_set)
        except ValueError:
            raise ValueError("Neighbors set not found in the structure environments")
        if nb_set_index == len(self.ce_list[isite][cn]):
            self.ce_list[isite][cn].append(ce)
        elif nb_set_index < len(self.ce_list[isite][cn]):
            self.ce_list[isite][cn][nb_set_index] = ce
        else:
            raise ValueError("Neighbors set not yet in ce_list !")

    def update_site_info(self, isite, info_dict):
        """
        Update information about this site.

        Args:
            isite: Index of the site for which info has to be updated.
            info_dict: Dictionary of information to be added for this site.
        """
        if "sites_info" not in self.info:
            self.info["sites_info"] = [{} for _ in range(len(self.structure))]
        self.info["sites_info"][isite].update(info_dict)

    def get_coordination_environments(self, isite, cn, nb_set):
        """
        Get the ChemicalEnvironments for a given site, coordination and neighbors set.

        Args:
            isite: Index of the site for which the ChemicalEnvironments is looked for.
            cn: Coordination for which the ChemicalEnvironments is looked for.
            nb_set: Neighbors set for which the ChemicalEnvironments is looked for.

        Returns:
            ChemicalEnvironments
        """
        if self.ce_list[isite] is None:
            return None
        if cn not in self.ce_list[isite]:
            return None
        try:
            nb_set_index = self.neighbors_sets[isite][cn].index(nb_set)
        except ValueError:
            return None
        return self.ce_list[isite][cn][nb_set_index]

    def get_csm(self, isite, mp_symbol):
        """
        Get the continuous symmetry measure for a given site in the given coordination environment.

        Args:
            isite: Index of the site.
            mp_symbol: Symbol of the coordination environment for which we want the continuous symmetry measure.

        Returns:
            Continuous symmetry measure of the given site in the given environment.
        """
        csms = self.get_csms(isite, mp_symbol)
        if len(csms) != 1:
            raise ChemenvError(
                "StructureEnvironments",
                "get_csm",
                f"Number of csms for site #{isite} with mp_symbol {mp_symbol!r} = {len(csms)}",
            )
        return csms[0]

    def get_csms(self, isite, mp_symbol) -> list:
        """
        Returns the continuous symmetry measure(s) of site with index isite with respect to the
        perfect coordination environment with mp_symbol. For some environments, a given mp_symbol might not
        be available (if there is no voronoi parameters leading to a number of neighbors corresponding to
        the coordination number of environment mp_symbol). For some environments, a given mp_symbol might
        lead to more than one csm (when two or more different voronoi parameters lead to different neighbors
        but with same number of neighbors).

        Args:
            isite: Index of the site.
            mp_symbol: MP symbol of the perfect environment for which the csm has to be given.

        Returns:
            list[CSM]: for site isite with respect to geometry mp_symbol
        """
        cn = symbol_cn_mapping[mp_symbol]
        if cn not in self.ce_list[isite]:
            return []
        return [envs[mp_symbol] for envs in self.ce_list[isite][cn]]

    def plot_csm_and_maps(self, isite, max_csm=8.0):
        """
        Plotting of the coordination numbers of a given site for all the distfactor/angfactor parameters. If the
        chemical environments are given, a color map is added to the plot, with the lowest continuous symmetry measure
        as the value for the color of that distfactor/angfactor set.

        Args:
            isite: Index of the site for which the plot has to be done
            max_csm: Maximum continuous symmetry measure to be shown.
        """
        fig = self.get_csm_and_maps(isite=isite, max_csm=max_csm)
        if fig is None:
            return
        plt.show()
        return

    def get_csm_and_maps(
        self, isite, max_csm=8.0, figsize=None, symmetry_measure_type=None
    ) -> tuple[plt.Figure, plt.Axes] | None:
        """
        Plotting of the coordination numbers of a given site for all the distfactor/angfactor parameters. If the
        chemical environments are given, a color map is added to the plot, with the lowest continuous symmetry measure
        as the value for the color of that distfactor/angfactor set.

        Args:
            isite: Index of the site for which the plot has to be done.
            max_csm: Maximum continuous symmetry measure to be shown.
            figsize: Size of the figure.
            symmetry_measure_type: Type of continuous symmetry measure to be used.

        Returns:
            Matplotlib figure and axes representing the CSM and maps.
        """
        if symmetry_measure_type is None:
            symmetry_measure_type = "csm_wcs_ctwcc"
        # Initializes the figure
        fig = plt.figure() if figsize is None else plt.figure(figsize=figsize)
        gs = GridSpec(2, 1, hspace=0.0, wspace=0.0)
        ax = fig.add_subplot(gs[:])
        ax_distang = ax.twinx()

        idx = 0
        cn_maps = []
        all_wds = []
        all_was = []
        max_wd = 0.0
        for cn, nb_sets in self.neighbors_sets[isite].items():
            for inb_set, nb_set in enumerate(nb_sets):
                ce = self.ce_list[isite][cn][inb_set]
                if ce is None:
                    continue
                min_geoms = ce.minimum_geometries(max_csm=max_csm)
                if len(min_geoms) == 0:
                    continue
                wds = nb_set.normalized_distances
                max_wd = max(max_wd, *wds)
                all_wds.append(wds)
                all_was.append(nb_set.normalized_angles)
                for mp_symbol, cg_dict in min_geoms:
                    csm = cg_dict["other_symmetry_measures"][symmetry_measure_type]
                    ax.plot(idx, csm, "ob")
                    ax.annotate(mp_symbol, xy=(idx, csm))
                cn_maps.append((cn, inb_set))
                idx += 1

        if max_wd < 1.225:
            ymax_wd = 1.25
            yticks_wd = np.linspace(1.0, ymax_wd, 6)
        elif max_wd < 1.36:
            ymax_wd = 1.4
            yticks_wd = np.linspace(1.0, ymax_wd, 5)
        elif max_wd < 1.45:
            ymax_wd = 1.5
            yticks_wd = np.linspace(1.0, ymax_wd, 6)
        elif max_wd < 1.55:
            ymax_wd = 1.6
            yticks_wd = np.linspace(1.0, ymax_wd, 7)
        elif max_wd < 1.75:
            ymax_wd = 1.8
            yticks_wd = np.linspace(1.0, ymax_wd, 5)
        elif max_wd < 1.95:
            ymax_wd = 2.0
            yticks_wd = np.linspace(1.0, ymax_wd, 6)
        elif max_wd < 2.35:
            ymax_wd = 2.5
            yticks_wd = np.linspace(1.0, ymax_wd, 7)
        else:
            ymax_wd = np.ceil(1.1 * max_wd)
            yticks_wd = np.linspace(1.0, ymax_wd, 6)

        yticks_wa = np.linspace(0.0, 1.0, 6)

        frac_bottom = 0.05
        frac_top = 0.05
        frac_middle = 0.1
        yamin = frac_bottom
        yamax = 0.5 - frac_middle / 2
        ydmin = 0.5 + frac_middle / 2
        ydmax = 1.0 - frac_top

        def yang(wa):
            return (yamax - yamin) * np.array(wa) + yamin

        def ydist(wd):
            return (np.array(wd) - 1.0) / (ymax_wd - 1.0) * (ydmax - ydmin) + ydmin

        for idx, was in enumerate(all_was):
            ax_distang.plot(0.2 + idx * np.ones_like(was), yang(was), "<g")
            alpha = 0.3 if np.mod(idx, 2) == 0 else 0.1
            ax_distang.fill_between(
                [-0.5 + idx, 0.5 + idx],
                [1.0, 1.0],
                0.0,
                facecolor="k",
                alpha=alpha,
                zorder=-1000,
            )
        for idx, wds in enumerate(all_wds):
            ax_distang.plot(0.2 + idx * np.ones_like(wds), ydist(wds), "sm")

        ax_distang.plot([-0.5, len(cn_maps)], [0.5, 0.5], "k--", alpha=0.5)

        yticks = yang(yticks_wa).tolist()
        yticks.extend(ydist(yticks_wd).tolist())
        yticklabels = yticks_wa.tolist()
        yticklabels.extend(yticks_wd.tolist())
        ax_distang.set_yticks(yticks)
        ax_distang.set_yticklabels(yticklabels)

        fake_subplot_ang = fig.add_subplot(gs[1], frame_on=False)
        fake_subplot_dist = fig.add_subplot(gs[0], frame_on=False)
        fake_subplot_ang.set_yticks([])
        fake_subplot_dist.set_yticks([])
        fake_subplot_ang.set_xticks([])
        fake_subplot_dist.set_xticks([])
        fake_subplot_ang.set_ylabel("Angle parameter", labelpad=45, rotation=-90)
        fake_subplot_dist.set_ylabel("Distance parameter", labelpad=45, rotation=-90)
        fake_subplot_ang.yaxis.set_label_position("right")
        fake_subplot_dist.yaxis.set_label_position("right")

        ax_distang.set_ylim([0.0, 1.0])
        ax.set_xticks(range(len(cn_maps)))
        ax.set_ylabel("Continuous symmetry measure")
        ax.set_xlim([-0.5, len(cn_maps) - 0.5])
        ax_distang.set_xlim([-0.5, len(cn_maps) - 0.5])
        ax.set_xticklabels([str(cn_map) for cn_map in cn_maps])

        return fig, ax

    def get_environments_figure(
        self,
        isite,
        plot_type=None,
        title="Coordination numbers",
        max_dist=2.0,
        colormap=None,
        figsize=None,
        strategy=None,
    ):
        """
        Plotting of the coordination environments of a given site for all the distfactor/angfactor regions. The
        chemical environments with the lowest continuous symmetry measure is shown for each distfactor/angfactor
        region as the value for the color of that distfactor/angfactor region (using a colormap).

        Args:
            isite: Index of the site for which the plot has to be done.
            plot_type: How to plot the coordinations.
            title: Title for the figure.
            max_dist: Maximum distance to be plotted when the plotting of the distance is set to 'initial_normalized'
                or 'initial_real' (Warning: this is not the same meaning in both cases! In the first case, the
                closest atom lies at a "normalized" distance of 1.0 so that 2.0 means refers to this normalized
                distance while in the second case, the real distance is used).
            colormap: Color map to be used for the continuous symmetry measure.
            figsize: Size of the figure.
            strategy: Whether to plot information about one of the Chemenv Strategies.

        Returns:
            tuple[plt.Figure, plt.Axes]: matplotlib figure and axes representing the environments.
        """
        # Initializes the figure
        fig = plt.figure(figsize=figsize)
        ax = fig.add_subplot(111)

        # Initializes the distance and angle parameters
        if plot_type is None:
            plot_type = {
                "distance_parameter": ("initial_normalized", None),
                "angle_parameter": ("initial_normalized_inverted", None),
            }

        clr_map = cm.jet if colormap is None else colormap
        clr_min = 0.0
        clr_max = 10.0
        norm = Normalize(vmin=clr_min, vmax=clr_max)
        scalarmap = cm.ScalarMappable(norm=norm, cmap=clr_map)
        dist_limits = [1.0, max_dist]
        ang_limits = [0.0, 1.0]
        if plot_type["distance_parameter"][0] == "one_minus_inverse_alpha_power_n":
            if plot_type["distance_parameter"][1] is None:
                exponent = 3
            else:
                exponent = plot_type["distance_parameter"][1]["exponent"]
            xlabel = f"Distance parameter : $1.0-\\frac{{1.0}}{{\\alpha^{{{exponent}}}}}$"

            def dp_func(dp):
                return 1.0 - 1.0 / np.power(dp, exponent)

        elif plot_type["distance_parameter"][0] == "initial_normalized":
            xlabel = "Distance parameter : $\\alpha$"

            def dp_func(dp):
                return dp

        else:
            raise ValueError(f"Wrong value for distance parameter plot type \"{plot_type['distance_parameter'][0]}\"")

        if plot_type["angle_parameter"][0] == "one_minus_gamma":
            ylabel = "Angle parameter : $1.0-\\gamma$"

            def ap_func(ap):
                return 1.0 - ap

        elif plot_type["angle_parameter"][0] in [
            "initial_normalized_inverted",
            "initial_normalized",
        ]:
            ylabel = "Angle parameter : $\\gamma$"

            def ap_func(ap):
                return ap

        else:
            raise ValueError(f"Wrong value for angle parameter plot type \"{plot_type['angle_parameter'][0]}\"")
        dist_limits = [dp_func(dp) for dp in dist_limits]
        ang_limits = [ap_func(ap) for ap in ang_limits]

        for cn, cn_nb_sets in self.neighbors_sets[isite].items():
            for inb_set, nb_set in enumerate(cn_nb_sets):
                nb_set_surface_pts = nb_set.voronoi_grid_surface_points()
                if nb_set_surface_pts is None:
                    continue
                ce = self.ce_list[isite][cn][inb_set]
                if ce is None:
                    color = "w"
                    inv_color = "k"
                    text = f"{cn}"
                else:
                    mingeom = ce.minimum_geometry()
                    if mingeom is not None:
                        mp_symbol = mingeom[0]
                        csm = mingeom[1]["symmetry_measure"]
                        color = scalarmap.to_rgba(csm)
                        inv_color = [
                            1.0 - color[0],
                            1.0 - color[1],
                            1.0 - color[2],
                            1.0,
                        ]
                        text = f"{mp_symbol}"
                    else:
                        color = "w"
                        inv_color = "k"
                        text = f"{cn}"
                nb_set_surface_pts = [(dp_func(pt[0]), ap_func(pt[1])) for pt in nb_set_surface_pts]
                polygon = Polygon(
                    nb_set_surface_pts,
                    closed=True,
                    edgecolor="k",
                    facecolor=color,
                    linewidth=1.2,
                )
                ax.add_patch(polygon)
                ipt = len(nb_set_surface_pts) / 2
                if ipt != int(ipt):
                    raise RuntimeError("Uneven number of surface points")
                ipt = int(ipt)
                patch_center = (
                    (nb_set_surface_pts[0][0] + min(nb_set_surface_pts[ipt][0], dist_limits[1])) / 2,
                    (nb_set_surface_pts[0][1] + nb_set_surface_pts[ipt][1]) / 2,
                )

                if (
                    np.abs(nb_set_surface_pts[-1][1] - nb_set_surface_pts[-2][1]) > 0.06
                    and np.abs(min(nb_set_surface_pts[-1][0], dist_limits[1]) - nb_set_surface_pts[0][0]) > 0.125
                ):
                    xytext = (
                        (min(nb_set_surface_pts[-1][0], dist_limits[1]) + nb_set_surface_pts[0][0]) / 2,
                        (nb_set_surface_pts[-1][1] + nb_set_surface_pts[-2][1]) / 2,
                    )
                    ax.annotate(
                        text,
                        xy=xytext,
                        ha="center",
                        va="center",
                        color=inv_color,
                        fontsize="x-small",
                    )
                elif (
                    np.abs(nb_set_surface_pts[ipt][1] - nb_set_surface_pts[0][1]) > 0.1
                    and np.abs(min(nb_set_surface_pts[ipt][0], dist_limits[1]) - nb_set_surface_pts[0][0]) > 0.125
                ):
                    xytext = patch_center
                    ax.annotate(
                        text,
                        xy=xytext,
                        ha="center",
                        va="center",
                        color=inv_color,
                        fontsize="x-small",
                    )

        ax.set_title(title)
        ax.set_xlabel(xlabel)
        ax.set_ylabel(ylabel)

        dist_limits.sort()
        ang_limits.sort()
        ax.set_xlim(dist_limits)
        ax.set_ylim(ang_limits)
        if strategy is not None:
            try:
                strategy.add_strategy_visualization_to_subplot(subplot=ax)
            except Exception:
                pass
        if plot_type["angle_parameter"][0] == "initial_normalized_inverted":
            ax.axes.invert_yaxis()

        scalarmap.set_array([clr_min, clr_max])
        cb = fig.colorbar(scalarmap, ax=ax, extend="max")
        cb.set_label("Continuous symmetry measure")
        return fig, ax

    def plot_environments(
        self,
        isite,
        plot_type=None,
        title="Coordination numbers",
        max_dist=2.0,
        figsize=None,
        strategy=None,
    ):
        """
        Plotting of the coordination numbers of a given site for all the distfactor/angfactor parameters. If the
        chemical environments are given, a color map is added to the plot, with the lowest continuous symmetry measure
        as the value for the color of that distfactor/angfactor set.

        Args:
            isite: Index of the site for which the plot has to be done.
            plot_type: How to plot the coordinations.
            title: Title for the figure.
            max_dist: Maximum distance to be plotted when the plotting of the distance is set to 'initial_normalized'
                or 'initial_real' (Warning: this is not the same meaning in both cases! In the first case, the
                closest atom lies at a "normalized" distance of 1.0 so that 2.0 means refers to this normalized
                distance while in the second case, the real distance is used).
            figsize: Size of the figure.
            strategy: Whether to plot information about one of the Chemenv Strategies.
        """
        fig, _ax = self.get_environments_figure(
            isite=isite,
            plot_type=plot_type,
            title=title,
            max_dist=max_dist,
            figsize=figsize,
            strategy=strategy,
        )
        if fig is None:
            return
        fig.show()

    def save_environments_figure(
        self,
        isite,
        imagename="image.png",
        plot_type=None,
        title="Coordination numbers",
        max_dist=2.0,
        figsize=None,
    ):
        """
        Saves the environments figure to a given file.

        Args:
            isite: Index of the site for which the plot has to be done.
            imagename: Name of the file to which the figure has to be saved.
            plot_type: How to plot the coordinations.
            title: Title for the figure.
            max_dist: Maximum distance to be plotted when the plotting of the distance is set to 'initial_normalized'
                or 'initial_real' (Warning: this is not the same meaning in both cases! In the first case, the
                closest atom lies at a "normalized" distance of 1.0 so that 2.0 means refers to this normalized
                distance while in the second case, the real distance is used).
            figsize: Size of the figure.
        """
        fig, _ax = self.get_environments_figure(
            isite=isite,
            plot_type=plot_type,
            title=title,
            max_dist=max_dist,
            figsize=figsize,
        )
        if fig is None:
            return
        fig.savefig(imagename)

    def differences_wrt(self, other):
        """
        Return differences found in the current StructureEnvironments with respect to another StructureEnvironments.

        Args:
            other: A StructureEnvironments object.

        Returns:
            List of differences between the two StructureEnvironments objects.
        """
        differences = []
        if self.structure != other.structure:
            differences += (
                {
                    "difference": "structure",
                    "comparison": "__eq__",
                    "self": self.structure,
                    "other": other.structure,
                },
                {"difference": "PREVIOUS DIFFERENCE IS DISMISSIVE", "comparison": "differences_wrt"},
            )
            return differences
        if self.valences != other.valences:
            differences.append(
                {
                    "difference": "valences",
                    "comparison": "__eq__",
                    "self": self.valences,
                    "other": other.valences,
                }
            )
        if self.info != other.info:
            differences.append(
                {
                    "difference": "info",
                    "comparison": "__eq__",
                    "self": self.info,
                    "other": other.info,
                }
            )
        if self.voronoi != other.voronoi:
            if self.voronoi.is_close_to(other.voronoi):
                differences += (
                    {"difference": "voronoi", "comparison": "__eq__", "self": self.voronoi, "other": other.voronoi},
                    {"difference": "PREVIOUS DIFFERENCE IS DISMISSIVE", "comparison": "differences_wrt"},
                )
                return differences

            differences += (
                {
                    "difference": "voronoi",
                    "comparison": "is_close_to",
                    "self": self.voronoi,
                    "other": other.voronoi,
                },
                {"difference": "PREVIOUS DIFFERENCE IS DISMISSIVE", "comparison": "differences_wrt"},
            )
            return differences
        for isite, self_site_nb_sets in enumerate(self.neighbors_sets):
            other_site_nb_sets = other.neighbors_sets[isite]
            if self_site_nb_sets is None:
                if other_site_nb_sets is None:
                    continue
                differences.append(
                    {
                        "difference": f"neighbors_sets[{isite=}]",
                        "comparison": "has_neighbors",
                        "self": "None",
                        "other": set(other_site_nb_sets),
                    }
                )
                continue
            if other_site_nb_sets is None:
                differences.append(
                    {
                        "difference": f"neighbors_sets[{isite=}]",
                        "comparison": "has_neighbors",
                        "self": set(self_site_nb_sets),
                        "other": "None",
                    }
                )
                continue
            self_site_cns = set(self_site_nb_sets)
            other_site_cns = set(other_site_nb_sets)
            if self_site_cns != other_site_cns:
                differences.append(
                    {
                        "difference": f"neighbors_sets[{isite=}]",
                        "comparison": "coordination_numbers",
                        "self": self_site_cns,
                        "other": other_site_cns,
                    }
                )
            common_cns = self_site_cns.intersection(other_site_cns)
            for cn in common_cns:
                other_site_cn_nb_sets = other_site_nb_sets[cn]
                self_site_cn_nb_sets = self_site_nb_sets[cn]
                set_self_site_cn_nb_sets = set(self_site_cn_nb_sets)
                set_other_site_cn_nb_sets = set(other_site_cn_nb_sets)
                if set_self_site_cn_nb_sets != set_other_site_cn_nb_sets:
                    differences.append(
                        {
                            "difference": f"neighbors_sets[{isite=}][{cn=}]",
                            "comparison": "neighbors_sets",
                            "self": self_site_cn_nb_sets,
                            "other": other_site_cn_nb_sets,
                        }
                    )
                common_nb_sets = set_self_site_cn_nb_sets.intersection(set_other_site_cn_nb_sets)
                for nb_set in common_nb_sets:
                    inb_set_self = self_site_cn_nb_sets.index(nb_set)
                    inb_set_other = other_site_cn_nb_sets.index(nb_set)
                    self_ce = self.ce_list[isite][cn][inb_set_self]
                    other_ce = other.ce_list[isite][cn][inb_set_other]
                    if self_ce != other_ce:
                        if self_ce.is_close_to(other_ce):
                            differences.append(
                                {
                                    "difference": f"ce_list[{isite=}][{cn=}][inb_set={inb_set_self}]",
                                    "comparison": "__eq__",
                                    "self": self_ce,
                                    "other": other_ce,
                                }
                            )
                        else:
                            differences.append(
                                {
                                    "difference": f"ce_list[{isite=}][{cn=}][inb_set={inb_set_self}]",
                                    "comparison": "is_close_to",
                                    "self": self_ce,
                                    "other": other_ce,
                                }
                            )
        return differences

    def __eq__(self, other: object) -> bool:
        if not isinstance(other, StructureEnvironments):
            return NotImplemented

        if len(self.ce_list) != len(other.ce_list):
            return False
        if self.voronoi != other.voronoi:
            return False
        if len(self.valences) != len(other.valences):
            return False
        if self.sites_map != other.sites_map:
            return False
        if self.equivalent_sites != other.equivalent_sites:
            return False
        if self.structure != other.structure:
            return False
        if self.info != other.info:
            return False
        for isite, site_ces in enumerate(self.ce_list):
            site_nb_sets_self = self.neighbors_sets[isite]
            site_nb_sets_other = other.neighbors_sets[isite]
            if site_nb_sets_self != site_nb_sets_other:
                return False
            if site_ces != other.ce_list[isite]:
                return False
        return True

    def as_dict(self):
        """
        Bson-serializable dict representation of the StructureEnvironments object.

        Returns:
            Bson-serializable dict representation of the StructureEnvironments object.
        """
        ce_list_dict = [
            {str(cn): [ce.as_dict() if ce is not None else None for ce in ce_dict[cn]] for cn in ce_dict}
            if ce_dict is not None
            else None
            for ce_dict in self.ce_list
        ]
        nbs_sets_dict = [
            {str(cn): [nb_set.as_dict() for nb_set in nb_sets] for cn, nb_sets in site_nbs_sets.items()}
            if site_nbs_sets is not None
            else None
            for site_nbs_sets in self.neighbors_sets
        ]
        info_dict = {key: val for key, val in self.info.items() if key not in ["sites_info"]}
        info_dict["sites_info"] = [
            {
                "nb_sets_info": {
                    str(cn): {str(inb_set): nb_set_info for inb_set, nb_set_info in cn_sets.items()}
                    for cn, cn_sets in site_info["nb_sets_info"].items()
                },
                "time": site_info["time"],
            }
            if "nb_sets_info" in site_info
            else {}
            for site_info in self.info["sites_info"]
        ]

        return {
            "@module": type(self).__module__,
            "@class": type(self).__name__,
            "voronoi": self.voronoi.as_dict(),
            "valences": self.valences,
            "sites_map": self.sites_map,
            "equivalent_sites": [[ps.as_dict() for ps in psl] for psl in self.equivalent_sites],
            "ce_list": ce_list_dict,
            "structure": self.structure.as_dict(),
            "neighbors_sets": nbs_sets_dict,
            "info": info_dict,
        }

    @classmethod
    def from_dict(cls, dct: dict) -> StructureEnvironments:
        """
        Reconstructs the StructureEnvironments object from a dict representation of the StructureEnvironments created
        using the as_dict method.

        Args:
            dct: dict representation of the StructureEnvironments object.

        Returns:
            StructureEnvironments object.
        """
        ce_list = [
            None
            if (ce_dict == "None" or ce_dict is None)
            else {
                int(cn): [
                    None if (ced is None or ced == "None") else ChemicalEnvironments.from_dict(ced)
                    for ced in ce_dict[cn]
                ]
                for cn in ce_dict
            }
            for ce_dict in dct["ce_list"]
        ]
        voronoi = DetailedVoronoiContainer.from_dict(dct["voronoi"])
        structure = Structure.from_dict(dct["structure"])
        neighbors_sets = [
            {
                int(cn): [
                    cls.NeighborsSet.from_dict(nb_set_dict, structure=structure, detailed_voronoi=voronoi)
                    for nb_set_dict in nb_sets
                ]
                for cn, nb_sets in site_nbs_sets_dict.items()
            }
            if site_nbs_sets_dict is not None
            else None
            for site_nbs_sets_dict in dct["neighbors_sets"]
        ]
        info = {key: val for key, val in dct["info"].items() if key not in ["sites_info"]}
        if "sites_info" in dct["info"]:
            info["sites_info"] = [
                {
                    "nb_sets_info": {
                        int(cn): {int(inb_set): nb_set_info for inb_set, nb_set_info in cn_sets.items()}
                        for cn, cn_sets in site_info["nb_sets_info"].items()
                    },
                    "time": site_info["time"],
                }
                if "nb_sets_info" in site_info
                else {}
                for site_info in dct["info"]["sites_info"]
            ]
        return cls(
            voronoi=voronoi,
            valences=dct["valences"],
            sites_map=dct["sites_map"],
            equivalent_sites=[[PeriodicSite.from_dict(psd) for psd in psl] for psl in dct["equivalent_sites"]],
            ce_list=ce_list,
            structure=structure,
            neighbors_sets=neighbors_sets,
            info=info,
        )


class LightStructureEnvironments(MSONable):
    """
    Class used to store the chemical environments of a given structure obtained from a given ChemenvStrategy. Currently,
    only strategies leading to the determination of a unique environment for each site is allowed
    This class does not store all the information contained in the StructureEnvironments object, only the coordination
    environment found.
    """

    DELTA_MAX_OXIDATION_STATE = 0.1
    DEFAULT_STATISTICS_FIELDS = (
        "anion_list",
        "anion_atom_list",
        "cation_list",
        "cation_atom_list",
        "neutral_list",
        "neutral_atom_list",
        "atom_coordination_environments_present",
        "ion_coordination_environments_present",
        "fraction_atom_coordination_environments_present",
        "fraction_ion_coordination_environments_present",
        "coordination_environments_atom_present",
        "coordination_environments_ion_present",
    )

    class NeighborsSet:
        """
        Class used to store a given set of neighbors of a given site (based on a list of sites, the voronoi
        container is not part of the LightStructureEnvironments object).
        """

        def __init__(self, structure: Structure, isite, all_nbs_sites, all_nbs_sites_indices):
            """Constructor for NeighborsSet.

            Args:
                structure: Structure object.
                isite: Index of the site for which neighbors are stored in this NeighborsSet.
                all_nbs_sites: All the possible neighbors for this site.
                all_nbs_sites_indices: Indices of the sites in all_nbs_sites that make up this NeighborsSet.
            """
            self.structure = structure
            self.isite = isite
            self.all_nbs_sites = all_nbs_sites
            indices = set(all_nbs_sites_indices)
            if len(indices) != len(all_nbs_sites_indices):
                raise ValueError("Set of neighbors contains duplicates !")
            self.all_nbs_sites_indices = sorted(indices)
            self.all_nbs_sites_indices_unsorted = all_nbs_sites_indices

        @property
        def neighb_coords(self):
            """Coordinates of neighbors for this NeighborsSet."""
            return [self.all_nbs_sites[inb]["site"].coords for inb in self.all_nbs_sites_indices_unsorted]

        @property
        def neighb_sites(self):
            """Neighbors for this NeighborsSet as pymatgen Sites."""
            return [self.all_nbs_sites[inb]["site"] for inb in self.all_nbs_sites_indices_unsorted]

        @property
        def neighb_sites_and_indices(self):
            """List of neighbors for this NeighborsSet as pymatgen Sites and their index in the original structure."""
            return [
                {
                    "site": self.all_nbs_sites[inb]["site"],
                    "index": self.all_nbs_sites[inb]["index"],
                }
                for inb in self.all_nbs_sites_indices_unsorted
            ]

        @property
        def neighb_indices_and_images(self) -> list[dict[str, int]]:
            """List of indices and images with respect to the original unit cell sites for this NeighborsSet."""
            return [
                {
                    "index": self.all_nbs_sites[inb]["index"],
                    "image_cell": self.all_nbs_sites[inb]["image_cell"],
                }
                for inb in self.all_nbs_sites_indices_unsorted
            ]

        def __len__(self) -> int:
            return len(self.all_nbs_sites_indices)

        def __hash__(self) -> int:
            return len(self.all_nbs_sites_indices)

        def __eq__(self, other: object) -> bool:
            needed_attrs = ("isite", "all_nbs_sites_indices")

            if not all(hasattr(other, attr) for attr in needed_attrs):
                return NotImplemented

            return all(getattr(self, attr) == getattr(other, attr) for attr in needed_attrs)

        def __str__(self):
            return (
                f"Neighbors Set for site #{self.isite} :\n"
                f" - Coordination number : {len(self)}\n"
                f" - Neighbors sites indices : {', '.join(f'{nb_idxs}' for nb_idxs in self.all_nbs_sites_indices)}\n"
            )

        def as_dict(self):
            """A JSON-serializable dict representation of the NeighborsSet."""
            return {
                "isite": self.isite,
                "all_nbs_sites_indices": self.all_nbs_sites_indices_unsorted,
            }

        @classmethod
        def from_dict(cls, dct, structure: Structure, all_nbs_sites):
            """
            Reconstructs the NeighborsSet algorithm from its JSON-serializable dict representation, together with
            the structure and all the possible neighbors sites.

            As an inner (nested) class, the NeighborsSet is not supposed to be used anywhere else that inside the
            LightStructureEnvironments. The from_dict method is thus using the structure and all_nbs_sites when
            reconstructing itself. These two are both in the LightStructureEnvironments object.

            Args:
                dct: a JSON-serializable dict representation of a NeighborsSet.
                structure: The structure.
                all_nbs_sites: The list of all the possible neighbors for a given site.

            Returns:
                NeighborsSet
            """
            return cls(
                structure=structure,
                isite=dct["isite"],
                all_nbs_sites=all_nbs_sites,
                all_nbs_sites_indices=dct["all_nbs_sites_indices"],
            )

    def __init__(
        self,
        strategy,
        coordination_environments=None,
        all_nbs_sites=None,
        neighbors_sets=None,
        structure=None,
        valences=None,
        valences_origin=None,
    ):
        """
        Constructor for the LightStructureEnvironments object.

        Args:
            strategy: ChemEnv strategy used to get the environments.
            coordination_environments: The coordination environments identified.
            all_nbs_sites: All the possible neighbors for each site in the structure.
            neighbors_sets: The neighbors sets of each site in the structure.
            structure: The structure.
            valences: The valences used to get the environments (if needed).
            valences_origin: How the valences were obtained (e.g. from the Bond-valence analysis or from the original
                structure).
        """
        self.strategy = strategy
        self.statistics_dict = None
        self.coordination_environments = coordination_environments
        self._all_nbs_sites = all_nbs_sites
        self.neighbors_sets = neighbors_sets
        self.structure = structure
        self.valences = valences
        self.valences_origin = valences_origin

    @classmethod
    def from_structure_environments(cls, strategy, structure_environments, valences=None, valences_origin=None):
        """
        Construct a LightStructureEnvironments object from a strategy and a StructureEnvironments object.

        Args:
            strategy: ChemEnv strategy used.
            structure_environments: StructureEnvironments object from which to construct the LightStructureEnvironments.
            valences: The valences of each site in the structure.
            valences_origin: How the valences were obtained (e.g. from the Bond-valence analysis or from the original
                structure).

        Returns:
            LightStructureEnvironments
        """
        structure = structure_environments.structure
        strategy.set_structure_environments(structure_environments=structure_environments)
        coordination_environments = [None] * len(structure)
        neighbors_sets = [None] * len(structure)
        _all_nbs_sites = []
        all_nbs_sites = []
        if valences is None:
            valences = structure_environments.valences
            if valences_origin is None:
                valences_origin = "from_structure_environments"
        elif valences_origin is None:
            valences_origin = "user-specified"

        for isite, site in enumerate(structure):
            site_ces_and_nbs_list = strategy.get_site_ce_fractions_and_neighbors(site, strategy_info=True)
            if site_ces_and_nbs_list is None:
                continue
            coordination_environments[isite] = []
            neighbors_sets[isite] = []
            site_ces = []
            site_nbs_sets = []
            for ce_and_neighbors in site_ces_and_nbs_list:
                _all_nbs_sites_indices = []
                # Coordination environment
                ce_dict = {
                    "ce_symbol": ce_and_neighbors["ce_symbol"],
                    "ce_fraction": ce_and_neighbors["ce_fraction"],
                }
                if ce_and_neighbors["ce_dict"] is not None:
                    csm = ce_and_neighbors["ce_dict"]["other_symmetry_measures"][strategy.symmetry_measure_type]
                else:
                    csm = None
                ce_dict["csm"] = csm
                ce_dict["permutation"] = (ce_and_neighbors.get("ce_dict") or {}).get("permutation")
                site_ces.append(ce_dict)
                # Neighbors
                neighbors = ce_and_neighbors["neighbors"]
                for nb_site_and_index in neighbors:
                    nb_site = nb_site_and_index["site"]
                    try:
                        nb_allnbs_sites_index = all_nbs_sites.index(nb_site)
                    except ValueError:
                        nb_index_unitcell = nb_site_and_index["index"]
                        diff = nb_site.frac_coords - structure[nb_index_unitcell].frac_coords
                        rounddiff = np.round(diff)
                        if not np.allclose(diff, rounddiff):
                            raise ValueError(
                                "Weird, differences between one site in a periodic image cell is not integer ..."
                            )
                        nb_image_cell = np.array(rounddiff, int)
                        nb_allnbs_sites_index = len(_all_nbs_sites)
                        _all_nbs_sites.append(
                            {"site": nb_site, "index": nb_index_unitcell, "image_cell": nb_image_cell}
                        )
                        all_nbs_sites.append(nb_site)
                    _all_nbs_sites_indices.append(nb_allnbs_sites_index)

                nb_set = cls.NeighborsSet(
                    structure=structure,
                    isite=isite,
                    all_nbs_sites=_all_nbs_sites,
                    all_nbs_sites_indices=_all_nbs_sites_indices,
                )
                site_nbs_sets.append(nb_set)
            coordination_environments[isite] = site_ces
            neighbors_sets[isite] = site_nbs_sets
        return cls(
            strategy=strategy,
            coordination_environments=coordination_environments,
            all_nbs_sites=_all_nbs_sites,
            neighbors_sets=neighbors_sets,
            structure=structure,
            valences=valences,
            valences_origin=valences_origin,
        )

    def setup_statistic_lists(self):
        """Set up the statistics of environments for this LightStructureEnvironments."""
        self.statistics_dict = {
            "valences_origin": self.valences_origin,
            "anion_list": {},  # OK
            "anion_number": None,  # OK
            "anion_atom_list": {},  # OK
            "anion_atom_number": None,  # OK
            "cation_list": {},  # OK
            "cation_number": None,  # OK
            "cation_atom_list": {},  # OK
            "cation_atom_number": None,  # OK
            "neutral_list": {},  # OK
            "neutral_number": None,  # OK
            "neutral_atom_list": {},  # OK
            "neutral_atom_number": None,  # OK
            "atom_coordination_environments_present": {},  # OK
            "ion_coordination_environments_present": {},  # OK
            "coordination_environments_ion_present": {},  # OK
            "coordination_environments_atom_present": {},  # OK
            "fraction_ion_coordination_environments_present": {},  # OK
            "fraction_atom_coordination_environments_present": {},  # OK
            "fraction_coordination_environments_ion_present": {},  # OK
            "fraction_coordination_environments_atom_present": {},  # OK
            "count_ion_present": {},  # OK
            "count_atom_present": {},  # OK
            "count_coordination_environments_present": {},
        }
        atom_stat = self.statistics_dict["atom_coordination_environments_present"]
        ce_atom_stat = self.statistics_dict["coordination_environments_atom_present"]
        fraction_atom_stat = self.statistics_dict["fraction_atom_coordination_environments_present"]
        fraction_ce_atom_stat = self.statistics_dict["fraction_coordination_environments_atom_present"]
        count_atoms = self.statistics_dict["count_atom_present"]
        count_ce = self.statistics_dict["count_coordination_environments_present"]
        for isite, site in enumerate(self.structure):
            # Building anion and cation list
            site_species = []
            if self.valences != "undefined":
                for sp, occ in site.species.items():
                    valence = self.valences[isite]
                    strspecie = str(Species(sp.symbol, valence))
                    if valence < 0:
                        specie_list = self.statistics_dict["anion_list"]
                        atomlist = self.statistics_dict["anion_atom_list"]
                    elif valence > 0:
                        specie_list = self.statistics_dict["cation_list"]
                        atomlist = self.statistics_dict["cation_atom_list"]
                    else:
                        specie_list = self.statistics_dict["neutral_list"]
                        atomlist = self.statistics_dict["neutral_atom_list"]
                    if strspecie not in specie_list:
                        specie_list[strspecie] = occ
                    else:
                        specie_list[strspecie] += occ
                    if sp.symbol not in atomlist:
                        atomlist[sp.symbol] = occ
                    else:
                        atomlist[sp.symbol] += occ
                    site_species.append((sp.symbol, valence, occ))
            # Building environments lists
            if self.coordination_environments[isite] is not None:
                site_envs = [
                    (ce_piece_dict["ce_symbol"], ce_piece_dict["ce_fraction"])
                    for ce_piece_dict in self.coordination_environments[isite]
                ]
                for ce_symbol, fraction in site_envs:
                    if fraction is None:
                        continue
                    count_ce.setdefault(ce_symbol, 0.0)
                    count_ce[ce_symbol] += fraction
                for sp, occ in site.species.items():
                    elmt = sp.symbol
                    if elmt not in atom_stat:
                        atom_stat[elmt] = {}
                        count_atoms[elmt] = 0.0
                    count_atoms[elmt] += occ
                    for ce_symbol, fraction in site_envs:
                        if fraction is None:
                            continue
                        if ce_symbol not in atom_stat[elmt]:
                            atom_stat[elmt][ce_symbol] = 0.0

                        atom_stat[elmt][ce_symbol] += occ * fraction
                        if ce_symbol not in ce_atom_stat:
                            ce_atom_stat[ce_symbol] = {}
                        if elmt not in ce_atom_stat[ce_symbol]:
                            ce_atom_stat[ce_symbol][elmt] = 0.0
                        ce_atom_stat[ce_symbol][elmt] += occ * fraction

                if self.valences != "undefined":
                    ion_stat = self.statistics_dict["ion_coordination_environments_present"]
                    ce_ion_stat = self.statistics_dict["coordination_environments_ion_present"]
                    count_ions = self.statistics_dict["count_ion_present"]
                    for elmt, oxi_state, occ in site_species:
                        if elmt not in ion_stat:
                            ion_stat[elmt] = {}
                            count_ions[elmt] = {}
                        if oxi_state not in ion_stat[elmt]:
                            ion_stat[elmt][oxi_state] = {}
                            count_ions[elmt][oxi_state] = 0.0
                        count_ions[elmt][oxi_state] += occ
                        for ce_symbol, fraction in site_envs:
                            if fraction is None:
                                continue
                            if ce_symbol not in ion_stat[elmt][oxi_state]:
                                ion_stat[elmt][oxi_state][ce_symbol] = 0.0
                            ion_stat[elmt][oxi_state][ce_symbol] += occ * fraction
                            if ce_symbol not in ce_ion_stat:
                                ce_ion_stat[ce_symbol] = {}
                            if elmt not in ce_ion_stat[ce_symbol]:
                                ce_ion_stat[ce_symbol][elmt] = {}
                            if oxi_state not in ce_ion_stat[ce_symbol][elmt]:
                                ce_ion_stat[ce_symbol][elmt][oxi_state] = 0.0
                            ce_ion_stat[ce_symbol][elmt][oxi_state] += occ * fraction
        self.statistics_dict["anion_number"] = len(self.statistics_dict["anion_list"])
        self.statistics_dict["anion_atom_number"] = len(self.statistics_dict["anion_atom_list"])
        self.statistics_dict["cation_number"] = len(self.statistics_dict["cation_list"])
        self.statistics_dict["cation_atom_number"] = len(self.statistics_dict["cation_atom_list"])
        self.statistics_dict["neutral_number"] = len(self.statistics_dict["neutral_list"])
        self.statistics_dict["neutral_atom_number"] = len(self.statistics_dict["neutral_atom_list"])

        for elmt, envs in atom_stat.items():
            sumelement = count_atoms[elmt]
            fraction_atom_stat[elmt] = {env: fraction / sumelement for env, fraction in envs.items()}
        for ce_symbol, atoms in ce_atom_stat.items():
            sumsymbol = count_ce[ce_symbol]
            fraction_ce_atom_stat[ce_symbol] = {atom: fraction / sumsymbol for atom, fraction in atoms.items()}
        ion_stat = self.statistics_dict["ion_coordination_environments_present"]
        fraction_ion_stat = self.statistics_dict["fraction_ion_coordination_environments_present"]
        ce_ion_stat = self.statistics_dict["coordination_environments_ion_present"]
        fraction_ce_ion_stat = self.statistics_dict["fraction_coordination_environments_ion_present"]
        count_ions = self.statistics_dict["count_ion_present"]
        for elmt, oxi_states_envs in ion_stat.items():
            fraction_ion_stat[elmt] = {}
            for oxi_state, envs in oxi_states_envs.items():
                sumspecie = count_ions[elmt][oxi_state]
                fraction_ion_stat[elmt][oxi_state] = {env: fraction / sumspecie for env, fraction in envs.items()}
        for ce_symbol, ions in ce_ion_stat.items():
            fraction_ce_ion_stat[ce_symbol] = {}
            sum_ce = np.sum([np.sum(list(oxi_states.values())) for elmt, oxi_states in ions.items()])
            for elmt, oxi_states in ions.items():
                fraction_ce_ion_stat[ce_symbol][elmt] = {
                    oxistate: fraction / sum_ce for oxistate, fraction in oxi_states.items()
                }

    def get_site_info_for_specie_ce(self, specie, ce_symbol):
        """
        Get list of indices that have the given specie with a given Coordination environment.

        Args:
            specie: Species to get.
            ce_symbol: Symbol of the coordination environment to get.

        Returns:
            dict: Keys are 'isites', 'fractions', 'csms' which contain list of indices in the structure
                that have the given specie in the given environment, their fraction and continuous
                symmetry measures.
        """
        element = specie.symbol
        oxi_state = specie.oxi_state
        isites = []
        csms = []
        fractions = []
        for isite, site in enumerate(self.structure):
            if element in [sp.symbol for sp in site.species] and (
                self.valences == "undefined" or oxi_state == self.valences[isite]
            ):
                for ce_dict in self.coordination_environments[isite]:
                    if ce_symbol == ce_dict["ce_symbol"]:
                        isites.append(isite)
                        csms.append(ce_dict["csm"])
                        fractions.append(ce_dict["ce_fraction"])
        return {"isites": isites, "fractions": fractions, "csms": csms}

    def get_site_info_for_specie_allces(self, specie, min_fraction=0):
        """
        Get list of indices that have the given specie.

        Args:
            specie: Species to get.
            min_fraction: Minimum fraction of the coordination environment.

        Returns:
            dict: with the list of coordination environments for the given species, the indices of the sites
                in which they appear, their fractions and continuous symmetry measures.
        """
        allces = {}
        element = specie.symbol
        oxi_state = specie.oxi_state
        for isite, site in enumerate(self.structure):
            if (
                element in [sp.symbol for sp in site.species] and self.valences == "undefined"
            ) or oxi_state == self.valences[isite]:
                if self.coordination_environments[isite] is None:
                    continue
                for ce_dict in self.coordination_environments[isite]:
                    if ce_dict["ce_fraction"] < min_fraction:
                        continue
                    if ce_dict["ce_symbol"] not in allces:
                        allces[ce_dict["ce_symbol"]] = {
                            "isites": [],
                            "fractions": [],
                            "csms": [],
                        }
                    allces[ce_dict["ce_symbol"]]["isites"].append(isite)
                    allces[ce_dict["ce_symbol"]]["fractions"].append(ce_dict["ce_fraction"])
                    allces[ce_dict["ce_symbol"]]["csms"].append(ce_dict["csm"])
        return allces

    def get_statistics(self, statistics_fields=DEFAULT_STATISTICS_FIELDS, bson_compatible=False):
        """
        Get the statistics of environments for this structure.

        Args:
            statistics_fields: Which statistics to get.
            bson_compatible: Whether to make the dictionary BSON-compatible.

        Returns:
            dict: with the requested statistics.
        """
        if self.statistics_dict is None:
            self.setup_statistic_lists()
        if statistics_fields == "ALL":
            statistics_fields = list(self.statistics_dict)
        if bson_compatible:
            dd = jsanitize({field: self.statistics_dict[field] for field in statistics_fields})
        else:
            dd = {field: self.statistics_dict[field] for field in statistics_fields}
        return dd

    def contains_only_one_anion_atom(self, anion_atom):
        """
        Whether this LightStructureEnvironments concerns a structure with only one given anion atom type.

        Args:
            anion_atom: Anion (e.g. O, ...). The structure could contain O2- and O- though.

        Returns:
            bool: True if this LightStructureEnvironments concerns a structure with only one given anion_atom.
        """
        return (
            len(self.statistics_dict["anion_atom_list"]) == 1 and anion_atom in self.statistics_dict["anion_atom_list"]
        )

    def contains_only_one_anion(self, anion):
        """
        Whether this LightStructureEnvironments concerns a structure with only one given anion type.

        Args:
            anion: Anion (e.g. O2-, ...).

        Returns:
            bool: True if this LightStructureEnvironments concerns a structure with only one given anion.
        """
        return len(self.statistics_dict["anion_list"]) == 1 and anion in self.statistics_dict["anion_list"]

    def site_contains_environment(self, isite, ce_symbol):
        """
        Whether a given site contains a given coordination environment.

        Args:
            isite: Index of the site.
            ce_symbol: Symbol of the coordination environment.

        Returns:
            bool: True if the site contains the given coordination environment.
        """
        if self.coordination_environments[isite] is None:
            return False
        return ce_symbol in [ce_dict["ce_symbol"] for ce_dict in self.coordination_environments[isite]]

    def site_has_clear_environment(self, isite, conditions=None):
        """
        Whether a given site has a "clear" environments.

        A "clear" environment is somewhat arbitrary. You can pass (multiple) conditions, e.g. the environment should
        have a continuous symmetry measure lower than this, a fraction higher than that, ...

        Args:
            isite: Index of the site.
            conditions: Conditions to be checked for an environment to be "clear".

        Returns:
            bool: True if the site has a clear environment.
        """
        if self.coordination_environments[isite] is None:
            raise ValueError(f"Coordination environments have not been determined for site {isite}")
        if conditions is None:
            return len(self.coordination_environments[isite]) == 1
        ce = max(self.coordination_environments[isite], key=lambda x: x["ce_fraction"])
        for condition in conditions:
            target = condition["target"]
            if target == "ce_fraction":
                if ce[target] < condition["minvalue"]:
                    return False
            elif target == "csm":
                if ce[target] > condition["maxvalue"]:
                    return False
            elif target == "number_of_ces":
                if ce[target] > condition["maxnumber"]:
                    return False
            else:
                raise ValueError(f"Target {target!r} for condition of clear environment is not allowed")
        return True

    def structure_has_clear_environments(self, conditions=None, skip_none=True, skip_empty=False):
        """
        Whether all sites in a structure have "clear" environments.

        Args:
            conditions: Conditions to be checked for an environment to be "clear".
            skip_none: Whether to skip sites for which no environments have been computed.
            skip_empty: Whether to skip sites for which no environments could be found.

        Returns:
            bool: True if all the sites in the structure have clear environments.
        """
        for isite in range(len(self.structure)):
            if self.coordination_environments[isite] is None:
                if skip_none:
                    continue
                return False
            if len(self.coordination_environments[isite]) == 0:
                if skip_empty:
                    continue
                return False
            if not self.site_has_clear_environment(isite=isite, conditions=conditions):
                return False
        return True

    def clear_environments(self, conditions=None):
        """
        Get the clear environments in the structure.

        Args:
            conditions: Conditions to be checked for an environment to be "clear".

        Returns:
            list: Clear environments in this structure.
        """
        clear_envs_list = set()
        for isite in range(len(self.structure)):
            if self.coordination_environments[isite] is None:
                continue
            if len(self.coordination_environments[isite]) == 0:
                continue
            if self.site_has_clear_environment(isite=isite, conditions=conditions):
                ce = max(
                    self.coordination_environments[isite],
                    key=lambda x: x["ce_fraction"],
                )
                clear_envs_list.add(ce["ce_symbol"])
        return list(clear_envs_list)

    def structure_contains_atom_environment(self, atom_symbol, ce_symbol):
        """
        Checks whether the structure contains a given atom in a given environment.

        Args:
            atom_symbol: Symbol of the atom.
            ce_symbol: Symbol of the coordination environment.

        Returns:
            bool: True if the coordination environment is found for the given atom.
        """
        for isite, site in enumerate(self.structure):
            if Element(atom_symbol) in site.species.element_composition and self.site_contains_environment(
                isite, ce_symbol
            ):
                return True
        return False

    def environments_identified(self):
        """
        Return the set of environments identified in this structure.

        Returns:
            set: environments identified in this structure.
        """
        return {ce["ce_symbol"] for celist in self.coordination_environments if celist is not None for ce in celist}

    @property
    def uniquely_determines_coordination_environments(self):
        """True if the coordination environments are uniquely determined."""
        return self.strategy.uniquely_determines_coordination_environments

    def __eq__(self, other: object) -> bool:
        """
        Equality method that checks if the LightStructureEnvironments object is equal to another
        LightStructureEnvironments object. Two LightStructureEnvironments objects are equal if the strategy used
        is the same, if the structure is the same, if the valences used in the strategies are the same, if the
        coordination environments and the neighbors determined by the strategy are the same.

        Args:
            other: LightStructureEnvironments object to compare with.

        Returns:
            bool: True if both objects are equal.
        """
        if not isinstance(other, LightStructureEnvironments):
            return NotImplemented

        is_equal = (
            self.strategy == other.strategy
            and self.structure == other.structure
            and self.coordination_environments == other.coordination_environments
            and self.valences == other.valences
            and self.neighbors_sets == other.neighbors_sets
        )
        this_sites = [ss["site"] for ss in self._all_nbs_sites]
        other_sites = [ss["site"] for ss in other._all_nbs_sites]
        this_indices = [ss["index"] for ss in self._all_nbs_sites]
        other_indices = [ss["index"] for ss in other._all_nbs_sites]
        return is_equal and this_sites == other_sites and this_indices == other_indices

    def as_dict(self):
        """
        Returns:
            dict: Bson-serializable representation of the LightStructureEnvironments object.
        """
        return {
            "@module": type(self).__module__,
            "@class": type(self).__name__,
            "strategy": self.strategy.as_dict(),
            "structure": self.structure.as_dict(),
            "coordination_environments": self.coordination_environments,
            "all_nbs_sites": [
                {
                    "site": PeriodicSite(
                        species=nb_site["site"].species,
                        coords=nb_site["site"].frac_coords,
                        lattice=nb_site["site"].lattice,
                        to_unit_cell=False,
                        coords_are_cartesian=False,
                        properties=nb_site["site"].properties,
                    ).as_dict(),
                    "index": nb_site["index"],
                    "image_cell": [int(ii) for ii in nb_site["image_cell"]],
                }
                for nb_site in self._all_nbs_sites
            ],
            "neighbors_sets": [
                [nb_set.as_dict() for nb_set in site_nb_sets] if site_nb_sets is not None else None
                for site_nb_sets in self.neighbors_sets
            ],
            "valences": self.valences,
        }

    @classmethod
    def from_dict(cls, dct) -> LightStructureEnvironments:
        """
        Reconstructs the LightStructureEnvironments object from a dict representation of the
        LightStructureEnvironments created using the as_dict method.

        Args:
            dct: dict representation of the LightStructureEnvironments object.

        Returns:
            LightStructureEnvironments object.
        """
        dec = MontyDecoder()
        structure = dec.process_decoded(dct["structure"])
        all_nbs_sites = []
        for nb_site in dct["all_nbs_sites"]:
            periodic_site = dec.process_decoded(nb_site["site"])
            site = PeriodicNeighbor(
                species=periodic_site.species,
                coords=periodic_site.frac_coords,
                lattice=periodic_site.lattice,
                properties=periodic_site.properties,
            )
            if "image_cell" in nb_site:
                image_cell = np.array(nb_site["image_cell"], int)
            else:
                diff = site.frac_coords - structure[nb_site["index"]].frac_coords
                rounddiff = np.round(diff)
                if not np.allclose(diff, rounddiff):
                    raise ValueError("Weird, differences between one site in a periodic image cell is not integer ...")
                image_cell = np.array(rounddiff, int)
            all_nbs_sites.append({"site": site, "index": nb_site["index"], "image_cell": image_cell})
        neighbors_sets = [
            [
                cls.NeighborsSet.from_dict(nb_set, structure=structure, all_nbs_sites=all_nbs_sites)
                for nb_set in site_nb_sets
            ]
            if site_nb_sets is not None
            else None
            for site_nb_sets in dct["neighbors_sets"]
        ]
        return cls(
            strategy=dec.process_decoded(dct["strategy"]),
            coordination_environments=dct["coordination_environments"],
            all_nbs_sites=all_nbs_sites,
            neighbors_sets=neighbors_sets,
            structure=structure,
            valences=dct["valences"],
        )


class ChemicalEnvironments(MSONable):
    """
    Class used to store all the information about the chemical environment of a given site for a given list of
    coordinated neighbors (internally called "cn_map").
    """

    def __init__(self, coord_geoms=None):
        """
        Initializes the ChemicalEnvironments object containing all the information about the chemical
        environment of a given site.

        Args:
            coord_geoms: coordination geometries to be added to the chemical environment.
        """
        if coord_geoms is None:
            self.coord_geoms = {}
        else:
            raise NotImplementedError(
                "Constructor for ChemicalEnvironments with the coord_geoms argument is not yet implemented"
            )

    def __getitem__(self, mp_symbol):
        return self.coord_geoms[mp_symbol]

    def __len__(self):
        """
        Returns the number of coordination geometries in this ChemicalEnvironments object.

        Returns:
            Number of coordination geometries in this ChemicalEnvironments object.
        """
        return len(self.coord_geoms)

    def __iter__(self):
        yield from self.coord_geoms.items()

    def minimum_geometry(self, symmetry_measure_type=None, max_csm=None):
        """
        Returns the geometry with the minimum continuous symmetry measure of this ChemicalEnvironments.

        Returns:
            tuple (symbol, csm) with symbol being the geometry with the minimum continuous symmetry measure and
            csm being the continuous symmetry measure associated to it.

        Raises:
            ValueError if no coordination geometry is found in this ChemicalEnvironments object.
        """
        if len(self.coord_geoms) == 0:
            return None
        cglist = list(self.coord_geoms)
        if symmetry_measure_type is None:
            csms = np.array([self.coord_geoms[cg]["other_symmetry_measures"]["csm_wcs_ctwcc"] for cg in cglist])
        else:
            csms = np.array([self.coord_geoms[cg]["other_symmetry_measures"][symmetry_measure_type] for cg in cglist])
        csmlist = [self.coord_geoms[cg] for cg in cglist]
        imin = np.argmin(csms)
        if max_csm is not None and csmlist[imin] > max_csm:
            return None
        return cglist[imin], csmlist[imin]

    def minimum_geometries(self, n=None, symmetry_measure_type=None, max_csm=None):
        """
        Returns a list of geometries with increasing continuous symmetry measure in this ChemicalEnvironments object.

        Args:
            n: Number of geometries to be included in the list.

        Returns:
            List of geometries with increasing continuous symmetry measure in this ChemicalEnvironments object.

        Raises:
            ValueError if no coordination geometry is found in this ChemicalEnvironments object.
        """
        cglist = list(self.coord_geoms)
        if symmetry_measure_type is None:
            csms = np.array([self.coord_geoms[cg]["other_symmetry_measures"]["csm_wcs_ctwcc"] for cg in cglist])
        else:
            csms = np.array([self.coord_geoms[cg]["other_symmetry_measures"][symmetry_measure_type] for cg in cglist])
        csmlist = [self.coord_geoms[cg] for cg in cglist]
        isorted = np.argsort(csms)
        if max_csm is not None:
            if n is None:
                return [(cglist[ii], csmlist[ii]) for ii in isorted if csms[ii] <= max_csm]

            return [(cglist[ii], csmlist[ii]) for ii in isorted[:n] if csms[ii] <= max_csm]

        if n is None:
            return [(cglist[ii], csmlist[ii]) for ii in isorted]
        return [(cglist[ii], csmlist[ii]) for ii in isorted[:n]]

    def add_coord_geom(
        self,
        mp_symbol,
        symmetry_measure,
        algo="UNKNOWN",
        permutation=None,
        override=False,
        local2perfect_map=None,
        perfect2local_map=None,
        detailed_voronoi_index=None,
        other_symmetry_measures=None,
        rotation_matrix=None,
        scaling_factor=None,
    ):
        """
        Adds a coordination geometry to the ChemicalEnvironments object.

        Args:
            mp_symbol: Symbol of the coordination geometry added.
            symmetry_measure: Symmetry measure of the coordination geometry added.
            algo: Algorithm used for the search of the coordination geometry added.
            permutation: Permutation of the neighbors that leads to the csm stored.
            override: If set to True, the coordination geometry will override the existent one if present.
            local2perfect_map: Mapping of the local indices to the perfect indices.
            perfect2local_map: Mapping of the perfect indices to the local indices.
            detailed_voronoi_index: Index in the voronoi containing the neighbors set.
            other_symmetry_measures: Other symmetry measure of the coordination geometry added (with/without the
                central atom, centered on the central atom or on the centroid with/without the central atom).
            rotation_matrix: Rotation matrix mapping the local geometry to the perfect geometry.
            scaling_factor: Scaling factor mapping the local geometry to the perfect geometry.

        Raises:
            ChemenvError if the coordination geometry is already added and override is set to False
        """
        if not all_cg.is_a_valid_coordination_geometry(mp_symbol=mp_symbol):
            raise ChemenvError(
                self.__class__,
                "add_coord_geom",
                f"Coordination geometry with mp_symbol {mp_symbol!r} is not valid",
            )
        if mp_symbol in list(self.coord_geoms) and not override:
            raise ChemenvError(
                self.__class__,
                "add_coord_geom",
                "This coordination geometry is already present and override is set to False",
            )

        self.coord_geoms[mp_symbol] = {
            "symmetry_measure": float(symmetry_measure),
            "algo": algo,
            "permutation": [int(i) for i in permutation],
            "local2perfect_map": local2perfect_map,
            "perfect2local_map": perfect2local_map,
            "detailed_voronoi_index": detailed_voronoi_index,
            "other_symmetry_measures": other_symmetry_measures,
            "rotation_matrix": rotation_matrix,
            "scaling_factor": scaling_factor,
        }

    def __str__(self):
        """
        Returns a string representation of the ChemicalEnvironments object.

        Returns:
            String representation of the ChemicalEnvironments object.
        """
        out = "Chemical environments object :\n"
        if len(self.coord_geoms) == 0:
            out += " => No coordination in it <=\n"
            return out
        for key in self.coord_geoms:
            mp_symbol = key
            break
        cn = symbol_cn_mapping[mp_symbol]
        out += f" => Coordination {cn} <=\n"
        mp_symbols = list(self.coord_geoms)
        csms_wcs = [self.coord_geoms[mp_symbol]["other_symmetry_measures"]["csm_wcs_ctwcc"] for mp_symbol in mp_symbols]
        icsms_sorted = np.argsort(csms_wcs)
        mp_symbols = [mp_symbols[ii] for ii in icsms_sorted]
        for mp_symbol in mp_symbols:
            csm_wcs = self.coord_geoms[mp_symbol]["other_symmetry_measures"]["csm_wcs_ctwcc"]
            csm_wocs = self.coord_geoms[mp_symbol]["other_symmetry_measures"]["csm_wocs_ctwocc"]
            out += f"   - {mp_symbol}\n"
            out += f"      csm1 (with central site) : {csm_wcs}"
            out += f"      csm2 (without central site) : {csm_wocs}"
            out += f"     algo : {self.coord_geoms[mp_symbol]['algo']}"
            out += f"     perm : {self.coord_geoms[mp_symbol]['permutation']}\n"
            out += f"       local2perfect : {self.coord_geoms[mp_symbol]['local2perfect_map']}\n"
            out += f"       perfect2local : {self.coord_geoms[mp_symbol]['perfect2local_map']}\n"
        return out

    def is_close_to(self, other, rtol=0.0, atol=1e-8) -> bool:
        """
        Whether this ChemicalEnvironments object is close to another one.

        Args:
            other: Another ChemicalEnvironments object.
            rtol: Relative tolerance for the comparison of Continuous Symmetry Measures.
            atol: Absolute tolerance for the comparison of Continuous Symmetry Measures.

        Returns:
            bool: True if the two ChemicalEnvironments objects are close to each other.
        """
        if set(self.coord_geoms) != set(other.coord_geoms):
            return False
        for mp_symbol, cg_dict_self in self.coord_geoms.items():
            cg_dict_other = other[mp_symbol]
            other_csms_self = cg_dict_self["other_symmetry_measures"]
            other_csms_other = cg_dict_other["other_symmetry_measures"]
            for csmtype in [
                "csm_wcs_ctwcc",
                "csm_wcs_ctwocc",
                "csm_wcs_csc",
                "csm_wocs_ctwcc",
                "csm_wocs_ctwocc",
                "csm_wocs_csc",
            ]:
                if not np.isclose(
                    other_csms_self[csmtype],
                    other_csms_other[csmtype],
                    rtol=rtol,
                    atol=atol,
                ):
                    return False
        return True

    def __eq__(self, other: object) -> bool:
        """
        Equality method that checks if the ChemicalEnvironments object is equal to another ChemicalEnvironments.
        object.

        Args:
            other: ChemicalEnvironments object to compare with.

        Returns:
            bool: True if both objects are equal.
        """
        if not isinstance(other, ChemicalEnvironments):
            return NotImplemented

        if set(self.coord_geoms) != set(other.coord_geoms):
            return False
        for mp_symbol, cg_dict_self in self.coord_geoms.items():
            cg_dict_other = other.coord_geoms[mp_symbol]
            if cg_dict_self["symmetry_measure"] != cg_dict_other["symmetry_measure"]:
                return False
            if cg_dict_self["algo"] != cg_dict_other["algo"]:
                return False
            if cg_dict_self["permutation"] != cg_dict_other["permutation"]:
                return False
            if cg_dict_self["detailed_voronoi_index"] != cg_dict_other["detailed_voronoi_index"]:
                return False
            other_csms_self = cg_dict_self["other_symmetry_measures"]
            other_csms_other = cg_dict_other["other_symmetry_measures"]
            for csmtype in [
                "csm_wcs_ctwcc",
                "csm_wcs_ctwocc",
                "csm_wcs_csc",
                "csm_wocs_ctwcc",
                "csm_wocs_ctwocc",
                "csm_wocs_csc",
            ]:
                if other_csms_self[csmtype] != other_csms_other[csmtype]:
                    return False
        return True

    def as_dict(self):
        """
        Returns a dictionary representation of the ChemicalEnvironments object.

        Returns:
            A dictionary representation of the ChemicalEnvironments object.
        """
        return {
            "@module": type(self).__module__,
            "@class": type(self).__name__,
            "coord_geoms": jsanitize(self.coord_geoms),
        }

    @classmethod
    def from_dict(cls, dct: dict) -> ChemicalEnvironments:
        """
        Reconstructs the ChemicalEnvironments object from a dict representation of the ChemicalEnvironments created
        using the as_dict method.

        Args:
            dct: dict representation of the ChemicalEnvironments object.

        Returns:
            ChemicalEnvironments object.
        """
        ce = cls()
        for cg in dct["coord_geoms"]:
            if dct["coord_geoms"][cg]["local2perfect_map"] is None:
                l2p_map = None
            else:
                l2p_map = {int(key): int(val) for key, val in dct["coord_geoms"][cg]["local2perfect_map"].items()}
            if dct["coord_geoms"][cg]["perfect2local_map"] is None:
                p2l_map = None
            else:
                p2l_map = {int(key): int(val) for key, val in dct["coord_geoms"][cg]["perfect2local_map"].items()}
            if (
                "other_symmetry_measures" in dct["coord_geoms"][cg]
                and dct["coord_geoms"][cg]["other_symmetry_measures"] is not None
            ):
                other_csms = dct["coord_geoms"][cg]["other_symmetry_measures"]
            else:
                other_csms = None
            ce.add_coord_geom(
                cg,
                dct["coord_geoms"][cg]["symmetry_measure"],
                dct["coord_geoms"][cg]["algo"],
                permutation=dct["coord_geoms"][cg]["permutation"],
                local2perfect_map=l2p_map,
                perfect2local_map=p2l_map,
                detailed_voronoi_index=dct["coord_geoms"][cg]["detailed_voronoi_index"],
                other_symmetry_measures=other_csms,
                rotation_matrix=dct["coord_geoms"][cg]["rotation_matrix"],
                scaling_factor=dct["coord_geoms"][cg]["scaling_factor"],
            )
        return ce