SingleCellExperiment 0.4.4__tar.gz → 0.4.6__tar.gz

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/.pre-commit-config.yaml

@@ -2,7 +2,7 @@ exclude: '^docs/conf.py'
 
 repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
-  rev: v4.5.0
+  rev: v4.6.0
   hooks:
   - id: trailing-whitespace
   - id: check-added-large-files
@@ -26,14 +26,14 @@ repos:
       # --config, ./pyproject.toml
 
 - repo: https://github.com/psf/black
-  rev: 24.1.1
+  rev: 24.4.0
   hooks:
   - id: black
     language_version: python3
 
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.2.1
+  rev: v0.3.7
   hooks:
     - id: ruff
       args: [--fix, --exit-non-zero-on-fix]

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: SingleCellExperiment
-Version: 0.4.4
+Version: 0.4.6
 Summary: Container class for single-cell experiments
 Home-page: https://github.com/BiocPy/SingleCellExperiment
 Author: jkanche
@@ -20,6 +20,7 @@ Requires-Dist: anndata; extra == "optional"
 Requires-Dist: h5py; extra == "optional"
 Requires-Dist: mudata; extra == "optional"
 Requires-Dist: delayedarray; extra == "optional"
+Requires-Dist: hdf5array; extra == "optional"
 Provides-Extra: testing
 Requires-Dist: setuptools; extra == "testing"
 Requires-Dist: pytest; extra == "testing"
@@ -28,6 +29,7 @@ Requires-Dist: anndata; extra == "testing"
 Requires-Dist: h5py; extra == "testing"
 Requires-Dist: mudata; extra == "testing"
 Requires-Dist: delayedarray; extra == "testing"
+Requires-Dist: hdf5array; extra == "testing"
 
 [![Project generated with PyScaffold](https://img.shields.io/badge/-PyScaffold-005CA0?logo=pyscaffold)](https://pyscaffold.org/)
 [![PyPI-Server](https://img.shields.io/pypi/v/SingleCellExperiment.svg)](https://pypi.org/project/SingleCellExperiment/)

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/setup.cfg

@@ -37,6 +37,7 @@ optional =
 	h5py
 	mudata
 	delayedarray
+	hdf5array
 testing = 
 	setuptools
 	pytest
@@ -45,6 +46,7 @@ testing =
 	h5py
 	mudata
 	delayedarray
+	hdf5array
 
 [options.entry_points]
 

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/src/SingleCellExperiment.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: SingleCellExperiment
-Version: 0.4.4
+Version: 0.4.6
 Summary: Container class for single-cell experiments
 Home-page: https://github.com/BiocPy/SingleCellExperiment
 Author: jkanche
@@ -20,6 +20,7 @@ Requires-Dist: anndata; extra == "optional"
 Requires-Dist: h5py; extra == "optional"
 Requires-Dist: mudata; extra == "optional"
 Requires-Dist: delayedarray; extra == "optional"
+Requires-Dist: hdf5array; extra == "optional"
 Provides-Extra: testing
 Requires-Dist: setuptools; extra == "testing"
 Requires-Dist: pytest; extra == "testing"
@@ -28,6 +29,7 @@ Requires-Dist: anndata; extra == "testing"
 Requires-Dist: h5py; extra == "testing"
 Requires-Dist: mudata; extra == "testing"
 Requires-Dist: delayedarray; extra == "testing"
+Requires-Dist: hdf5array; extra == "testing"
 
 [![Project generated with PyScaffold](https://img.shields.io/badge/-PyScaffold-005CA0?logo=pyscaffold)](https://pyscaffold.org/)
 [![PyPI-Server](https://img.shields.io/pypi/v/SingleCellExperiment.svg)](https://pypi.org/project/SingleCellExperiment/)

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/src/SingleCellExperiment.egg-info/requires.txt

@@ -8,6 +8,7 @@ anndata
 h5py
 mudata
 delayedarray
+hdf5array
 
 [testing]
 setuptools
@@ -17,3 +18,4 @@ anndata
 h5py
 mudata
 delayedarray
+hdf5array

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/src/singlecellexperiment/SingleCellExperiment.py

@@ -1015,7 +1015,9 @@ class SingleCellExperiment(RangedSummarizedExperiment):
                 Input data.
 
         Returns:
-            A ``SingleCellExperiment`` object.
+            A ``SingleCellExperiment`` object. If the input contains any data
+            in the ``uns`` attribute, the `metadata` slot of the ``SingleCellExperiment``
+            will contain a key ``uns``.
         """
 
         layers = OrderedDict()
@@ -1028,12 +1030,13 @@ class SingleCellExperiment(RangedSummarizedExperiment):
         obsm = _to_normal_dict(input.obsm)
         varp = _to_normal_dict(input.varp)
         obsp = _to_normal_dict(input.obsp)
+        _metadata = {"uns": input.uns} if input.uns is not None else None
 
         return cls(
             assays=layers,
             row_data=biocframe.BiocFrame.from_pandas(input.var),
             column_data=biocframe.BiocFrame.from_pandas(input.obs),
-            metadata=input.uns,
+            metadata=_metadata,
             reduced_dims=obsm,
             row_pairs=varp,
             column_pairs=obsp,

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/src/singlecellexperiment/io/tenx.py

@@ -44,7 +44,7 @@ def read_tenx_mtx(path: str) -> SingleCellExperiment:
     )
 
 
-def read_tenx_h5(path: str) -> SingleCellExperiment:
+def read_tenx_h5(path: str, realize_assays: bool = False) -> SingleCellExperiment:
     """Read 10X H5 file as :py:class:`~singlecellexperiment.SingleCellExperiment.SingleCellExperiment`.
 
     Note: Currently only supports version 3 of the 10X H5 format.
@@ -53,12 +53,17 @@ def read_tenx_h5(path: str) -> SingleCellExperiment:
         path:
             Path to 10x H5 file.
 
+        realize_assays:
+            Whether to realize assays into memory.
+            Defaults to False.
+
     Returns:
         A single-cell experiment object.
     """
 
+    import delayedarray as da
     import h5py
-    from scipy.sparse import csc_matrix, csr_matrix
+    from hdf5array import Hdf5CompressedSparseMatrix
 
     h5 = h5py.File(path, mode="r")
 
@@ -68,16 +73,14 @@ def read_tenx_h5(path: str) -> SingleCellExperiment:
     groups = h5["matrix"].keys()
 
     # read the matrix
-    data = h5["matrix"]["data"][:]
-    indices = h5["matrix"]["indices"][:]
-    indptr = h5["matrix"]["indptr"][:]
     shape = tuple(h5["matrix"]["shape"][:])
 
-    counts = None
-    if len(indptr) == shape[1] + 1:
-        counts = csc_matrix((data, indices, indptr), shape=shape)
-    else:
-        counts = csr_matrix((data, indices, indptr), shape=shape)
+    counts = Hdf5CompressedSparseMatrix(
+        path=path, group_name="matrix", by_column=True, shape=shape
+    )
+
+    if realize_assays is True:
+        counts = da.to_scipy_sparse_matrix(counts, "csr")
 
     # read features
     features = None
@@ -106,6 +109,8 @@ def read_tenx_h5(path: str) -> SingleCellExperiment:
         barcodes["barcodes"] = [x.decode("ascii") for x in h5["matrix"]["barcodes"]]
         barcodes = BiocFrame(barcodes, number_of_rows=counts.shape[1])
 
+    h5.close()
+
     return SingleCellExperiment(
         assays={"counts": counts}, row_data=features, column_data=barcodes
     )

{SingleCellExperiment-0.4.4 → singlecellexperiment-0.4.6}/tests/test_sce_io.py

@@ -6,9 +6,11 @@ from biocframe import BiocFrame
 import numpy as np
 import pandas as pd
 from mudata import MuData
+from scipy import sparse
 
 import singlecellexperiment
 from singlecellexperiment.SingleCellExperiment import SingleCellExperiment
+from hdf5array import Hdf5CompressedSparseMatrix
 
 __author__ = "jkanche"
 __copyright__ = "jkanche"
@@ -95,6 +97,7 @@ def test_SCE_from10xH5():
     assert isinstance(tse, SingleCellExperiment)
 
     assert tse.assays is not None
+    assert isinstance(tse.assay(0), Hdf5CompressedSparseMatrix)
     assert tse.row_data is not None
     assert tse.col_data is not None
 
@@ -110,6 +113,11 @@ def test_SCE_from10xH5():
 
     assert sliced.shape == (10, 4)
 
+    tse = singlecellexperiment.read_tenx_h5(
+        "tests/data/tenx.sub.h5", realize_assays=True
+    )
+    assert isinstance(tse.assay(0), sparse.spmatrix)
+
 
 def test_SCE_randomAnnData():
     np.random.seed(1)
@@ -129,6 +137,14 @@ def test_SCE_randomAnnData():
     assert tse is not None
     assert isinstance(tse, SingleCellExperiment)
 
+    # to avoid unknown mapping types;
+    # ran into an issue with anndata.compat._overloaded_dict.OverloadedDict when loading a h5ad
+    adata.uns = {".internal": [f"obs_{i+1}" for i in range(n)]}
+    tse = singlecellexperiment.SingleCellExperiment.from_anndata(adata)
+
+    assert tse is not None
+    assert isinstance(tse, SingleCellExperiment)
+
 
 def test_SCE_to_mudata():
     tse = SingleCellExperiment(