fix bug when running no normalization and handle pipeline without pca

src/nichepca/workflows/_nichepca.py

@@ -11,7 +11,7 @@
     resolve_graph_constructor,
 )
 from nichepca.nhood_embedding import aggregate
-from nichepca.utils import check_for_raw_counts, normalize_per_sample
+from nichepca.utils import check_for_raw_counts, normalize_per_sample, to_numpy
 
 if TYPE_CHECKING:
     from anndata import AnnData
@@ -74,19 +74,22 @@ def nichepca(
     -------
     None
     """
+    # make sure pipeline is an iterable
+    if isinstance(pipeline, str):
+        pipeline = [pipeline]
+
     # we always need to use agg
     assert "agg" in pipeline, "aggregation must be part of the pipeline"
+
     # assert that the pca is behind norm and log1p
-    if "pca" in pipeline:
-        pca_after_norm = np.argmax(np.array(pipeline) == "pca") > np.argmax(
-            np.array(pipeline) == "norm"
-        )
-        pca_after_log1p = np.argmax(np.array(pipeline) == "pca") > np.argmax(
-            np.array(pipeline) == "log1p"
-        )
+    if "pca" in pipeline and ("norm" in pipeline or "log1p" in pipeline):
+        pca_index = np.argmax(np.array(pipeline) == "pca")
+        norm_index = np.argmax(np.array(pipeline) == "norm")
+        log1p_index = np.argmax(np.array(pipeline) == "log1p")
+        # argmax returns 0 if not found
         assert (
-            pca_after_norm and pca_after_log1p
-        ), "pca must be executed after norm and log1p"
+            norm_index <= pca_index and log1p_index <= pca_index
+        ), "PCA must be executed after both norm and log1p."
 
     # perform sanity check in case we are normalizing the data
     if "norm" or "log1p" in pipeline and obs_key is None and obsm_key is None:
@@ -170,14 +173,17 @@ def nichepca(
     # extract the results and remove old keys
     if "X_pca_harmony" in ad_tmp.obsm:
         X_npca = ad_tmp.obsm["X_pca_harmony"]
-    else:
+    elif "X_pca" in ad_tmp.obsm:
         X_npca = ad_tmp.obsm["X_pca"]
+    else:
+        X_npca = to_numpy(ad_tmp.X)
 
     # store the results
     adata.obsm["X_npca"] = X_npca
-    adata.uns["npca"] = ad_tmp.uns["pca"]
-    adata.uns["npca"]["PCs"] = pd.DataFrame(
-        data=ad_tmp.varm["PCs"],
-        index=ad_tmp.var_names,
-        columns=[f"PC{i}" for i in range(n_comps)],
-    )
+    if "pca" in pipeline:
+        adata.uns["npca"] = ad_tmp.uns["pca"]
+        adata.uns["npca"]["PCs"] = pd.DataFrame(
+            data=ad_tmp.varm["PCs"],
+            index=ad_tmp.var_names,
+            columns=[f"PC{i}" for i in range(n_comps)],
+        )

tests/test_workflows.py

@@ -54,6 +54,20 @@ def test_nichepca_single():
 
     assert np.all(adata_1.obsm["X_npca"] == adata_2.obsm["X_pca"])
 
+    # test without pca
+    pipeline = "agg"
+
+    adata = generate_dummy_adata()
+    npc.wf.nichepca(adata, knn=5, pipeline=pipeline)
+    X_npca_0 = adata.obsm["X_npca"]
+
+    adata = generate_dummy_adata()
+    npc.gc.knn_graph(adata, knn=5)
+    npc.ne.aggregate(adata)
+    X_npca_1 = npc.utils.to_numpy(adata.X)
+
+    assert np.all(X_npca_0 == X_npca_1)
+
 
 def test_nichepca_multi_sample():
     adata_1 = generate_dummy_adata()