Backport PR #3045 on branch 1.2.x (feat(external): adding batch_key a…

…nd labels_key) (#3049) Backport PR #3045: feat(external): adding batch_key and labels_key Co-authored-by: Yishen Miao <[email protected]>
scverse · Nov 19, 2024 · 8ad0139 · 8ad0139
1 parent c189c28
commit 8ad0139
Show file tree

Hide file tree

Showing 3 changed files with 17 additions and 8 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -14,12 +14,14 @@ to [Semantic Versioning]. Full commit history is available in the
     `datasplitter_kwargs={"drop_last": False}` and `train_size = None` by moving them into
     validation set, if available.
     {pr}`3036`.
+- Add `batch_key` and `labels_key` to `scvi.external.SCAR.setup_anndata`.
 
 #### Fixed
 
 - Breaking Change: Fix `get_outlier_cell_sample_pairs` function in {class}`scvi.external.MRVI`
     to correctly compute the maxmimum log-density across in-sample cells rather than the
     aggregated posterior log-density {pr}`3007`.
+- Fix references to `scvi.external` in `scvi.external.SCAR.setup_anndata`.
 
 #### Changed
 

diff --git a/src/scvi/external/scar/_model.py b/src/scvi/external/scar/_model.py
@@ -40,7 +40,7 @@ class SCAR(RNASeqMixin, VAEMixin, UnsupervisedTrainingMixin, BaseModelClass):
     Parameters
     ----------
     adata
-        AnnData object that has been registered via :meth:`~scvi_external.SCAR.setup_anndata`.
+        AnnData object that has been registered via :meth:`~scvi.external.SCAR.setup_anndata`.
     ambient_profile
         The probability of occurrence of each ambient transcript.\
             If None, averaging cells to estimate the ambient profile, by default None.
@@ -70,15 +70,15 @@ class SCAR(RNASeqMixin, VAEMixin, UnsupervisedTrainingMixin, BaseModelClass):
         the sparsity should be low; on the other hand, it should be set high
         in the case of unflitered genes.
     **model_kwargs
-        Keyword args for :class:`~scvi_external.SCAR`
+        Keyword args for :class:`~scvi.external.SCAR`
 
     Examples
     --------
     >>> adata = anndata.read_h5ad(path_to_anndata)
     >>> raw_adata = anndata.read_h5ad(path_to_raw_anndata)
-    >>> scvi_external.SCAR.setup_anndata(adata, batch_key="batch")
-    >>> scvi_external.SCAR.get_ambient_profile(adata=adata, raw_adata=raw_adata, prob=0.995)
-    >>> vae = scvi_external.SCAR(adata)
+    >>> scvi.external.SCAR.setup_anndata(adata, batch_key="batch")
+    >>> scvi.external.SCAR.get_ambient_profile(adata=adata, raw_adata=raw_adata, prob=0.995)
+    >>> vae = scvi.external.SCAR(adata)
     >>> vae.train()
     >>> adata.obsm["X_scAR"] = vae.get_latent_representation()
     >>> adata.layers['denoised'] = vae.get_denoised_counts()
@@ -152,6 +152,8 @@ def __init__(
     def setup_anndata(
         cls,
         adata: AnnData,
+        batch_key: str | None = None,
+        labels_key: str | None = None,
         layer: str | None = None,
         size_factor_key: str | None = None,
         **kwargs,
@@ -161,14 +163,16 @@ def setup_anndata(
         Parameters
         ----------
         %(param_adata)s
+        %(param_batch_key)s
+        %(param_labels_key)s
         %(param_layer)s
         %(param_size_factor_key)s
         """
         setup_method_args = cls._get_setup_method_args(**locals())
         anndata_fields = [
             LayerField(REGISTRY_KEYS.X_KEY, layer, is_count_data=True),
-            CategoricalObsField(REGISTRY_KEYS.BATCH_KEY, None),
-            CategoricalObsField(REGISTRY_KEYS.LABELS_KEY, None),
+            CategoricalObsField(REGISTRY_KEYS.BATCH_KEY, batch_key),
+            CategoricalObsField(REGISTRY_KEYS.LABELS_KEY, labels_key),
             NumericalObsField(REGISTRY_KEYS.SIZE_FACTOR_KEY, size_factor_key, required=False),
         ]
         adata_manager = AnnDataManager(fields=anndata_fields, setup_method_args=setup_method_args)

diff --git a/tests/external/scar/test_scar.py b/tests/external/scar/test_scar.py
@@ -7,8 +7,11 @@
 def test_scar():
     n_latent = 5
     adata = synthetic_iid()
+
+    adata.obs["batch"] = adata.obs["batch"].cat.rename_categories(["batch_2", "batch_3"])
+
     adata.X = scipy.sparse.csr_matrix(adata.X)
-    SCAR.setup_anndata(adata)
+    SCAR.setup_anndata(adata, batch_key="batch", labels_key="labels")
 
     _ = SCAR.get_ambient_profile(adata, adata, prob=0.0, iterations=1, sample=100)
     model = SCAR(adata, ambient_profile=None, n_latent=n_latent)