Cleaned up definition of emulsions, allowing to define empty emulsions

* Removed `grid` attribute from `Emulsion` and similar classes
* Instead, grid instances can now be specified when distances are
important, which separates concerns better
* Added explicit `dtype` attribute to emulsions to clearly specify type
of droplets in empty emulsions. This can often be ignored, though.

droplets/droplet_tracks.py

@@ -254,7 +254,7 @@ def time_overlaps(self, other: "DropletTrack") -> bool:
         return s0 <= o1 and o0 <= s1
 
     @classmethod
-    def _from_hdf_dataset(cls, dataset) -> "DropletTrack":
+    def _from_hdf_dataset(cls, dataset) -> DropletTrack:
         """construct a droplet track by reading data from an hdf5 dataset
 
         Args:
@@ -455,7 +455,9 @@ def __getitem__(self, key: Union[int, slice]):  # type: ignore
     def from_emulsion_time_course(
         cls,
         time_course: EmulsionTimeCourse,
+        *,
         method: str = "overlap",
+        grid: Optional[GridBase] = None,
         progress: bool = False,
         **kwargs,
     ) -> DropletTrackList:
@@ -469,6 +471,9 @@ def from_emulsion_time_course(
                 "overlap" (adding droplets that overlap with those in previous frames)
                 and "distance" (matching droplets to minimize center-to-center
                 distances).
+            grid (:class:`~pde.grids.base.GridBase`):
+                The grid on which the droplets are defined, which is necessary if
+                periodic boundary conditions should be respected for measuring distances
             progress (bool):
                 Whether to show the progress of the process.
             **kwargs:
@@ -496,7 +501,7 @@ def match_tracks(
                     # determine which old tracks could be extended
                     overlaps: List[DropletTrack] = []
                     for track in tracks_alive:
-                        if track.last.overlaps(droplet, time_course.grid):
+                        if track.last.overlaps(droplet, grid=grid):
                             overlaps.append(track)
 
                     if len(overlaps) == 1:
@@ -522,10 +527,10 @@ def match_tracks(
 
                 # calculate the distance between droplets
                 if tracks_alive:
-                    if time_course.grid is None:
+                    if grid is None:
                         metric: Union[str, Callable] = "euclidean"
                     else:
-                        metric = time_course.grid.distance_real
+                        metric = grid.distance_real
                     points_prev = [track.last.position for track in tracks_alive]
                     points_now = [droplet.position for droplet in emulsion]
                     dists = distance.cdist(points_prev, points_now, metric=metric)

droplets/droplets.py

@@ -1,6 +1,9 @@
 """
 Classes representing (perturbed) droplets in various dimensions
 
+The classes differ in what features of a droplet they track. In the simplest case, only
+the position and radius of a spherical droplet are stored. Other classes additionally
+keep track of the interfacial width or shape perturbations (of various degrees).
 
 .. autosummary::
    :nosignatures:
@@ -240,7 +243,7 @@ def data_bounds(self) -> Tuple[np.ndarray, np.ndarray]:
         return np.full(num, -np.inf), np.full(num, np.inf)
 
 
-class SphericalDroplet(DropletBase):  # lgtm [py/missing-equals]
+class SphericalDroplet(DropletBase):
     """Represents a single, spherical droplet"""
 
     __slots__ = ["data"]
@@ -1181,13 +1184,14 @@ def _get_mpl_patch(self, dim=None, *, color=None, **kwargs):
         raise NotImplementedError("Plotting PerturbedDroplet3D is not implemented")
 
 
-def droplet_from_data(droplet_class: str, data) -> DropletBase:
+def droplet_from_data(droplet_class: str, data: np.ndarray) -> DropletBase:
     """create a droplet instance of the given class using some data
 
     Args:
-        droplet_class (str): The name of the class that is used to create the
-            droplet instance
-        data (numpy.ndarray): A numpy array that defines the droplet properties
+        droplet_class (str):
+            The name of the class that is used to create the droplet instance
+        data (:class:`~numpy.ndarray`):
+            A numpy array that defines the droplet properties
     """
     cls = DropletBase._subclasses[droplet_class]
     return cls(**{key: data[key] for key in data.dtype.names})  # type: ignore