🎨 Introduce ordered collections and simplify .mapped()

lamindb/_collection.py

@@ -312,7 +312,7 @@ def from_artifacts(artifacts: Iterable[Artifact]) -> Tuple[str, Dict[str, str]]:
 def mapped(
     self,
     label_keys: Optional[Union[str, List[str]]] = None,
-    join_vars: Optional[Literal["auto", "inner", "outer"]] = "auto",
+    join: Optional[Literal["inner", "outer"]] = "outer",
     encode_labels: bool = True,
     cache_categories: bool = True,
     parallel: bool = False,
@@ -333,7 +333,7 @@ def mapped(
     return MappedCollection(
         path_list,
         label_keys,
-        join_vars,
+        join,
         encode_labels,
         cache_categories,
         parallel,

lamindb/dev/_mapped_collection.py

@@ -58,13 +58,13 @@ def __init__(
         self,
         path_list: List[Union[str, PathLike]],
         label_keys: Optional[Union[str, List[str]]] = None,
-        join_vars: Optional[Literal["auto", "inner", "outer"]] = "auto",
+        join: Optional[Literal["inner", "outer"]] = "outer",
         encode_labels: bool = True,
         cache_categories: bool = True,
         parallel: bool = False,
         dtype: Optional[str] = None,
     ):
-        assert join_vars in {None, "auto", "inner", "outer"}
+        assert join in {None, "inner", "outer"}
 
         self.storages = []  # type: ignore
         self.conns = []  # type: ignore
@@ -84,7 +84,7 @@ def __init__(
         self.indices = np.hstack([np.arange(n_obs) for n_obs in self.n_obs_list])
         self.storage_idx = np.repeat(np.arange(len(self.storages)), self.n_obs_list)
 
-        self.join_vars = join_vars if len(path_list) > 1 else None
+        self.join_vars = join if len(path_list) > 1 else None
         self.var_indices = None
         if self.join_vars is not None:
             self._make_join_vars()
@@ -100,7 +100,6 @@ def __init__(
                 self._make_encoders(self.label_keys)
 
         self._dtype = dtype
-
         self._closed = False
 
     def _make_connections(self, path_list: list, parallel: bool):

tests/test_collection.py

@@ -300,9 +300,7 @@ def test_collection_mapped():
     assert ls_ds.closed
     del ls_ds
 
-    with collection.mapped(
-        label_keys="feat1", join_vars="inner", dtype="float32"
-    ) as ls_ds:
+    with collection.mapped(label_keys="feat1", join="inner", dtype="float32") as ls_ds:
         assert not ls_ds.closed
         assert len(ls_ds) == 4
         assert len(ls_ds[0]) == 2 and len(ls_ds[2]) == 2
@@ -314,8 +312,8 @@ def test_collection_mapped():
     assert len(ls_ds[0]) == 2 and len(ls_ds[2]) == 2
 
     # adata3 goes first here
-    with collection_outer.mapped(label_keys="feat1", join_vars="auto") as ls_ds:
-        assert ls_ds.join_vars == "outer"
+    with collection_outer.mapped(label_keys="feat1", join="inner") as ls_ds:
+        assert ls_ds.join_vars == "inner"
         assert len(ls_ds.var_joint) == 6
         assert len(ls_ds[0]) == 2
         assert len(ls_ds[0][0]) == 6