aggregate the anndata object (#100)

* move utils to get

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* switch utils to get

* fix for docs

* release notes

* added kernels

* added aggregate templates

* fixed releasenotes

* update kernel

* updates _mean_var_major

* updated kernel

* updated Aggregate

* updates docs

* updates docstrings

* removes small

* adds scanpy-1.10 as dep

* updated test

* added dense_kernel

* changed utils test

* remove singledispatch

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

docs/Usage_Principles.md

@@ -23,11 +23,11 @@ adata.X = cpx.scipy.sparse.csr_matrix(adata.X)  # moves `.X` to the GPU
 adata.X = adata.X.get() # moves `.X` back to the CPU
 ```
 
-You can also use {mod}`rapids_singlecell.utils` to move arrays and matrices.
+You can also use {mod}`rapids_singlecell.get` to move arrays and matrices.
 
 ```
-rsc.utils.anndata_to_GPU(adata) # moves `.X` to the GPU
-rsc.utils.anndata_to_CPU(adata) # moves `.X` to the CPU
+rsc.get.anndata_to_GPU(adata) # moves `.X` to the GPU
+rsc.get.anndata_to_CPU(adata) # moves `.X` to the CPU
 ```
 
 ### Preprocessing

docs/api/utils.md → docs/api/get.md

@@ -1,14 +1,15 @@
-# Utils
+# Get object from `AnnData`: `get`
 
 These functions offer convineant ways to move arrays and matrices from and to the GPU.
 
 ```{eval-rst}
-.. module:: rapids_singlecell.utils
+.. module:: rapids_singlecell.get
 .. currentmodule:: rapids_singlecell
 
 .. autosummary::
     :toctree: generated
 
-    utils.anndata_to_GPU
-    utils.anndata_to_CPU
+    get.aggregate
+    get.anndata_to_GPU
+    get.anndata_to_CPU
 ```

docs/api/index.md

@@ -13,6 +13,6 @@ import rapids_singlecell as rsc
 scanpy_gpu
 squidpy_gpu
 decoupler_gpu
-utils
+get
 cunndata
 ```

docs/notebooks/demo_gpu-PR.ipynb

@@ -120,7 +120,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_GPU(adata)"
+    "rsc.get.anndata_to_GPU(adata)"
    ]
   },
   {
@@ -735,7 +735,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_CPU(adata, convert_all=True)"
+    "rsc.get.anndata_to_CPU(adata, convert_all=True)"
    ]
   },
   {

docs/notebooks/demo_gpu-seuratv3-brain-1M.ipynb

@@ -140,7 +140,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_GPU(adata)"
+    "rsc.get.anndata_to_GPU(adata)"
    ]
   },
   {
@@ -630,7 +630,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_CPU(adata, layer=\"counts\")"
+    "rsc.get.anndata_to_CPU(adata, layer=\"counts\")"
    ]
   },
   {
@@ -868,7 +868,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_CPU(adata)"
+    "rsc.get.anndata_to_CPU(adata)"
    ]
   },
   {

docs/notebooks/demo_gpu-seuratv3.ipynb

@@ -125,7 +125,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_GPU(adata)"
+    "rsc.get.anndata_to_GPU(adata)"
    ]
   },
   {
@@ -859,7 +859,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_CPU(adata, convert_all=True)"
+    "rsc.get.anndata_to_CPU(adata, convert_all=True)"
    ]
   },
   {

docs/notebooks/demo_gpu.ipynb

@@ -119,7 +119,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_GPU(adata)"
+    "rsc.get.anndata_to_GPU(adata)"
    ]
   },
   {
@@ -659,7 +659,7 @@
    ],
    "source": [
     "%%time\n",
-    "rsc.utils.anndata_to_CPU(adata)"
+    "rsc.get.anndata_to_CPU(adata)"
    ]
   },
   {

docs/release-notes/0.10.0.md

@@ -0,0 +1,8 @@
+### 0.10.0 {small}`the future`
+
+```{rubric} Features
+```
+
+* added `get` module {pr}`100` {smaller}`S Dicks`
+* switch `utils` functions to get {pr}`100` {smaller}`S Dicks`
+* added `get.aggregated` to create condensed `anndata` objects {pr}`100` {smaller}`S Dicks`

docs/release-notes/0.9.0.md

@@ -5,7 +5,7 @@
 * added support for GPU enabled {class}`~anndata.AnnData` {pr}`60` {smaller}`S Dicks`
 * added support for dense `cupy.ndarray` {pr}`60` {smaller}`S Dicks`
 * added support for 64 bit floats {pr}`60` {smaller}`S Dicks`
-* added {mod}`~rapids_singlecell.utils` to help with {class}`~anndata.AnnData` GPU conversion  {pr}`60` {smaller}`S Dicks`
+* added `rapids_singlecell.utils` to help with {class}`~anndata.AnnData` GPU conversion  {pr}`60` {smaller}`S Dicks`
 * renamed `cunndata_funcs` to {mod}`~rapids_singlecell.pp` {pr}`60` {smaller}`S Dicks`
 * renamed `scanpy_gpu` to {mod}`~rapids_singlecell.tl` {pr}`60` {smaller}`S Dicks`
 * switched {func}`~rapids_singlecell.pp.neighbors` to {mod}`~rapids_singlecell.pp` {pr}`60` {smaller}`S Dicks`

docs/release-notes/0.9.6.md

@@ -1,4 +1,4 @@
-### 0.9.6 {small}
+### 0.9.6
 
 ```{rubric} Bug fixes
 ```

docs/release-notes/index.md

@@ -2,6 +2,11 @@
 
 # Release notes
 
+## Version 0.10.0
+
+```{include} /release-notes/0.10.0.md
+``````
+
 ## Version 0.9.0
 ```{include} /release-notes/0.9.6.md
 ``````

pyproject.toml

@@ -13,7 +13,7 @@ dynamic = ["version"]
 
 dependencies = [
     "anndata>=0.10.0",
-    "scanpy>=1.9.3",
+    "scanpy>=1.10.0rc2",
     "numpy>=1.17.0",
     "decoupler>=1.3.2",
     "scipy>=1.4",
@@ -82,6 +82,7 @@ ignore = [
 "docs/*" = ["I"]
 "tests/*" = ["D"]
 "*/__init__.py" = ["F401"]
+"src/rapids_singlecell/get/_aggregated.py" = ["F811"]
 
 [tool.hatch.build]
 # exclude big files that don’t need to be installed

src/rapids_singlecell/__init__.py

@@ -1,2 +1,2 @@
-from . import cunnData, dcg, gr, pp, tl, utils
+from . import cunnData, dcg, get, gr, pp, tl
 from ._version import __version__

src/rapids_singlecell/get/__init__.py

@@ -0,0 +1,2 @@
+from ._aggregated import aggregate
+from ._anndata import anndata_to_CPU, anndata_to_GPU