Merge pull request #136 from scverse/fix-plus

scanpy 1.10 compat + PR CI

.github/workflows/pythonpackage.yml

@@ -1,37 +1,45 @@
 name: Python package
 
-on: [push]
+on:
+  push:
+    branches: [master]
+  pull_request:
+    branches: [master]
+  schedule:
+    - cron: "0 5 1,15 * *"
 
 jobs:
   build:
 
     runs-on: ubuntu-latest
     strategy:
+      fail-fast: false
       matrix:
-        python-version: [3.8, 3.9]
+        python-version: [3.8, 3.12]
 
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4
     - name: Set up Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v1
+      uses: actions/setup-python@v5
       with:
         python-version: ${{ matrix.python-version }}
+    - name: Install tools
+      run: python -m pip install --upgrade pip
     - name: Install dependencies
+      run: >
+        pip install
+        ".[test]"
+        git+https://github.com/bioFAM/mofapy2
+        git+https://github.com/scverse/mudata
+    - name: List dependencies
       run: |
-        python -m pip install --upgrade pip
-        python -m pip install umap-learn
-        python -m pip install .
-        # python -m pip install mofapy2
-        python -m pip install git+https://github.com/bioFAM/mofapy2
-        python -m pip install git+https://github.com/scverse/mudata
+        pip list
     - name: Lint with flake8
       run: |
-        pip install flake8
         # stop the build if there are Python syntax errors or undefined names
         flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
         # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
         flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
     - name: Test with pytest
       run: |
-        pip install pytest
         pytest

docs/source/changelog.rst

@@ -9,6 +9,12 @@ Release notes
 
    *
 
+
+v0.1.6
+------
+
+* Compatibility with scanpy 1.10
+
 v0.1.5
 ------
 

muon/__init__.py

@@ -13,4 +13,4 @@
 from . import atac
 from . import prot
 
-__version__ = "0.1.5"
+__version__ = "0.1.6"

muon/_core/preproc.py

@@ -18,14 +18,21 @@
 from sklearn.utils import check_random_state
 
 from anndata import AnnData
+import scanpy
 from scanpy import logging
 from scanpy.tools._utils import _choose_representation
-from scanpy.neighbors import _compute_connectivities_umap
 from umap.distances import euclidean
 from umap.sparse import sparse_euclidean, sparse_jaccard
 from umap.umap_ import nearest_neighbors
 from numba import njit, prange
 
+from packaging.version import Version
+
+if Version(scanpy.__version__) < Version("1.10"):
+    from scanpy.neighbors import _compute_connectivities_umap
+else:
+    from scanpy.neighbors._connectivity import umap as _compute_connectivities_umap
+
 from mudata import MuData
 
 # Computational methods for preprocessing
@@ -162,7 +169,7 @@ def _make_slice_intervals(idx, maxsize=10000):
 def _l2norm(
     adata: AnnData, rep: Optional[Union[Iterable[str], str]] = None, n_pcs: Optional[int] = 0
 ):
-    X = _choose_representation(adata, rep, n_pcs)
+    X = _choose_representation(adata=adata, use_rep=rep, n_pcs=n_pcs)
     sparse_X = issparse(X)
     if sparse_X:
         X_norm = linalg.norm(X, ord=2, axis=1)
@@ -211,7 +218,7 @@ def l2norm(
             rep = next(it)
             try:
                 next(it)
-            except StopIteration as e:
+            except StopIteration:
                 pass
             else:
                 raise RuntimeError("If 'rep' is an Iterable, it must have length 1")
@@ -220,7 +227,7 @@ def l2norm(
             n_pcs = next(it)
             try:
                 next(it)
-            except StopIteration as e:
+            except StopIteration:
                 pass
             else:
                 raise RuntimeError("If 'n_pcs' is an Iterable, it must have length 1")
@@ -358,7 +365,7 @@ def neighbors(
         mod_neighbors[i] = nparams["params"].get("n_neighbors", 0)
 
         neighbors_params[mod] = nparams
-        reps[mod] = _choose_representation(mdata.mod[mod], use_rep, n_pcs)
+        reps[mod] = _choose_representation(adata=mdata.mod[mod], use_rep=use_rep, n_pcs=n_pcs)
         mod_reps[mod] = (
             use_rep if use_rep is not None else -1
         )  # otherwise this is not saved to h5mu
@@ -585,7 +592,7 @@ def neighdist(cell, nz):
     neighbordistances = _sparse_csr_fast_knn(neighbordistances, n_neighbors + 1)
 
     logging.info("Calculating connectivities...")
-    _, connectivities = _compute_connectivities_umap(
+    connectivities = _compute_connectivities_umap(
         knn_indices=neighbordistances.indices.reshape(
             (neighbordistances.shape[0], n_neighbors + 1)
         ),
@@ -599,8 +606,8 @@ def neighdist(cell, nz):
         conns_key = "connectivities"
         dists_key = "distances"
     else:
-        conns_key = key_added + "_connectivities"
-        dists_key = key_added + "_distances"
+        conns_key = f"{key_added}_connectivities"
+        dists_key = f"{key_added}_distances"
     neighbors_dict = {"connectivities_key": conns_key, "distances_key": dists_key}
     neighbors_dict["params"] = {
         "n_neighbors": n_neighbors,
@@ -711,7 +718,7 @@ def func(x):
             else:
                 obs_subset = data.obs_names.isin(var)
         else:
-            raise ValueError(f"When providing obs_names directly, func has to be None.")
+            raise ValueError("When providing obs_names directly, func has to be None.")
 
     # Subset .obs
     data._obs = data.obs[obs_subset]
@@ -819,12 +826,9 @@ def func(x):
             )
     else:
         if func is None:
-            if np.array(var).dtype == bool:
-                var_subset = var
-            else:
-                var_subset = data.var_names.isin(var)
+            var_subset = var if np.array(var).dtype == bool else data.var_names.isin(var)
         else:
-            raise ValueError(f"When providing var_names directly, func has to be None.")
+            raise ValueError("When providing var_names directly, func has to be None.")
 
     # Subset .var
     data._var = data.var[var_subset]

muon/_core/tools.py

@@ -17,7 +17,8 @@
 
 from scanpy import logging
 from scanpy.tools._utils import _choose_representation
-from scanpy.neighbors import _compute_connectivities_umap
+
+# from scanpy.neighbors import _compute_connectivities_umap
 
 from typing import Union, Optional, List, Iterable, Mapping, Sequence, Type, Any, Dict, Literal
 from types import MappingProxyType
@@ -435,11 +436,9 @@ def mofa(
     if outfile is None:
         outfile = os.path.join("/tmp", "mofa_{}.hdf5".format(strftime("%Y%m%d-%H%M%S")))
 
-    if use_var:
-        if use_var not in data.var.columns:
-            warn(f"There is no column {use_var} in the provided object")
-            use_var = None
-
+    if use_var and use_var not in data.var.columns:
+        warn(f"There is no column {use_var} in the provided object")
+        use_var = None
     if isinstance(data, MuData):
         common_obs = reduce(np.intersect1d, [v.obs_names.values for k, v in mdata.mod.items()])
         if len(common_obs) != mdata.n_obs:
@@ -457,9 +456,8 @@ def mofa(
     ent = entry_point()
 
     lik = likelihoods
-    if lik is not None:
-        if isinstance(lik, str) and isinstance(lik, Iterable):
-            lik = [lik for _ in range(len(mdata.mod))]
+    if lik is not None and (isinstance(lik, str) and isinstance(lik, Iterable)):
+        lik = [lik for _ in range(len(mdata.mod))]
 
     ent.set_data_options(
         scale_views=scale_views,
@@ -787,7 +785,7 @@ def snf(
         mod_neighbors[i] = nparams["params"].get("n_neighbors", 0)
 
         neighbors_params[mod] = nparams
-        reps[mod] = _choose_representation(mdata.mod[mod], use_rep, n_pcs)
+        reps[mod] = _choose_representation(adata=mdata.mod[mod], use_rep=use_rep, n_pcs=n_pcs)
         mod_reps[mod] = (
             use_rep if use_rep is not None else -1
         )  # otherwise this is not saved to h5mu
@@ -855,7 +853,7 @@ def _normalize(x):
     def _dominateset(x, k=20):
         def _zero(arr):
             if k >= len(arr):
-                raise ValueError(f"'n_neighbors' seems to be too high.")
+                raise ValueError("'n_neighbors' seems to be too high.")
             arr = arr.copy()
             arr[np.argsort(arr)[: (len(arr) - k)]] = 0
             return arr
@@ -1319,7 +1317,7 @@ def umap(
     n_pcs = {k: (v if v != -1 else None) for k, v in nparams["n_pcs"].items()}
     observations = mdata.obs.index
     for mod, rep in use_rep.items():
-        rep = _choose_representation(mdata.mod[mod], rep, n_pcs[mod])
+        rep = _choose_representation(adata=mdata.mod[mod], use_rep=rep, n_pcs=n_pcs[mod])
         nfeatures += rep.shape[1]
         reps[mod] = rep
     rep = np.empty((len(observations), nfeatures), np.float32)

pyproject.toml

@@ -46,6 +46,10 @@ atac = [
     "pybedtools",
     "pysam"
 ]
+test = [
+    "pytest",
+    "flake8",
+]
 
 [tool.flit.metadata.urls]
 Documentation = "https://muon.readthedocs.io/en/latest/"