Remove the confusing expansion of feature intervals (#18)

CellArr · Jan 30, 2025 · e9e0eb6 · e9e0eb6
1 parent d3cc814
commit e9e0eb6
Show file tree

Hide file tree

Showing 4 changed files with 33 additions and 27 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,8 +1,9 @@
 # Changelog
 
-## Version 0.2.1
+## Version 0.2.1 - 0.2.2
 
 - The aggregate function is expected to return either a scalar value or a 1-dimensional NumPy ndarray. If the later, users need to specify the expected dimension of the summarization. All values will be flattenned eventually.
+- Remove expanding the intervals to conform to output length; this is now incompatible with coercions to anndata and summarized experiments and has been removed.
 
 ## Version 0.2.0
 

diff --git a/src/genomicarrays/GenomicArrayDataset.py b/src/genomicarrays/GenomicArrayDataset.py
@@ -287,22 +287,23 @@ def get_slice(
 
         if not isinstance(feature_subset, (int, slice)):
             raise TypeError("feature indices must be continous; either a 'slice' or 'int' index.")
+
         _fsubset = self.get_feature_subset(feature_subset)
         start_findex = _fsubset["genarr_feature_start_index"].astype(int).min()
         end_findex = _fsubset["genarr_feature_end_index"].astype(int).max()
 
         # expand intervals
-        final_rows = []
-        for row in _fsubset.itertuples():
-            for i, _ in enumerate(range(int(row.genarr_feature_start_index), int(row.genarr_feature_end_index))):
-                final_rows.append(row._replace(starts=i + row.starts, ends=i + row.starts + 1))
-        _feature_df = pd.DataFrame(final_rows)
+        # final_rows = []
+        # for row in _fsubset.itertuples():
+        #     for i, _ in enumerate(range(int(row.genarr_feature_start_index), int(row.genarr_feature_end_index))):
+        #         final_rows.append(row._replace(starts=i + row.starts, ends=i + row.starts + 1))
+        # _feature_df = pd.DataFrame(final_rows)
 
         _msubset = self.get_matrix_subset((list(range(start_findex, end_findex)), _sample_indices))
 
         return GenomicArrayDatasetSlice(
             _ssubset,
-            _feature_df,
+            _fsubset,
             _msubset,
         )
 

diff --git a/src/genomicarrays/GenomicArrayDatasetSlice.py b/src/genomicarrays/GenomicArrayDatasetSlice.py
@@ -51,22 +51,22 @@ class GenomicArrayDatasetSlice:
     ## Interop
     ####
 
-    def to_anndata(self):
-        """Convert the realized slice to :py:class:`~anndata.AnnData`."""
-        return anndata.AnnData(
-            layers={"matrix": self.matrix.transpose()},
-            obs=self.sample_metadata,
-            var=self.feature_annotation,
-        )
-
-    def to_rangedsummarizedexperiment(self):
-        """Convert the realized slice to
-        :py:class:`~summarizedexperiment.RangedSummarizedExperiment.RangedSummarizedExperiment`."""
-        return se.RangedSummarizedExperiment(
-            assays={"matrix": self.matrix},
-            row_ranges=gr.GenomicRanges.from_pandas(self.feature_annotation),
-            column_data=self.sample_metadata,
-        )
+    # def to_anndata(self):
+    #     """Convert the realized slice to :py:class:`~anndata.AnnData`."""
+    #     return anndata.AnnData(
+    #         layers={"matrix": self.matrix.transpose()},
+    #         obs=self.sample_metadata,
+    #         var=self.feature_annotation,
+    #     )
+
+    # def to_rangedsummarizedexperiment(self):
+    #     """Convert the realized slice to
+    #     :py:class:`~summarizedexperiment.RangedSummarizedExperiment.RangedSummarizedExperiment`."""
+    #     return se.RangedSummarizedExperiment(
+    #         assays={"matrix": self.matrix},
+    #         row_ranges=gr.GenomicRanges.from_pandas(self.feature_annotation),
+    #         column_data=self.sample_metadata,
+    #     )
 
     ####
     ## Misc methods.

diff --git a/tests/test_query.py b/tests/test_query.py
@@ -93,10 +93,12 @@ def test_query_agg():
 
     result1 = cd.get_slice(slice(0, 5), slice(None))
     assert result1 is not None
+    assert len(result1.feature_annotation) == 6
+    assert len(result1.sample_metadata) == 2
     assert result1.matrix.shape == (6, 2)
 
-    assert result1.to_anndata() is not None
-    assert result1.to_rangedsummarizedexperiment() is not None
+    # assert result1.to_anndata() is not None
+    # assert result1.to_rangedsummarizedexperiment() is not None
 
 
 def test_query_noagg():
@@ -165,7 +167,9 @@ def test_query_noagg():
 
     result1 = cd.get_slice(slice(0, 5), slice(None))
     assert result1 is not None
+    assert len(result1.feature_annotation) == 6
+    assert len(result1.sample_metadata) == 2
     assert result1.matrix.shape == (90, 2)
 
-    assert result1.to_anndata() is not None
-    assert result1.to_rangedsummarizedexperiment() is not None
+    # assert result1.to_anndata() is not None
+    # assert result1.to_rangedsummarizedexperiment() is not None