Remove dask-ml dependency (since sgkit now includes TruncatedSVD)

Author: tomwhite

.github/scripts/upstream_install.py

@@ -14,9 +14,8 @@ def install_deps() -> None:
         "--upgrade",
     )
     upstream_deps = (
-        "git+https://github.com/dask/dask.git#egg=dask[array]",
+        "git+https://github.com/dask/dask.git#egg=dask[array,dataframe]",
         "git+https://github.com/dask/distributed.git#egg=distributed",
-        "git+https://github.com/dask/dask-ml.git#egg=dask-ml",
         "git+https://github.com/pandas-dev/pandas#egg=pandas",
         "git+https://github.com/pangeo-data/rechunker.git#egg=rechunker",
         "git+https://github.com/pydata/xarray.git#egg=xarray",

requirements-numpy2.txt

@@ -1,8 +1,7 @@
 numpy < 2.1
 xarray
-dask[array] >= 2023.01.0, <= 2024.8.0
+dask[array,dataframe] >= 2023.01.0, <= 2024.8.0
 distributed >= 2023.01.0, <= 2024.8.0
-dask-ml
 scipy
 typing-extensions
 numba

requirements.txt

@@ -1,8 +1,7 @@
 numpy < 2
 xarray
-dask[array] >= 2023.01.0, <= 2024.8.0
+dask[array,dataframe] >= 2023.01.0, <= 2024.8.0
 distributed >= 2023.01.0, <= 2024.8.0
-dask-ml
 scipy
 typing-extensions
 numba

setup.cfg

@@ -29,9 +29,8 @@ python_requires = >=3.10
 install_requires =
     numpy < 2
     xarray
-    dask[array] >= 2022.01.0, <= 2024.8.0
+    dask[array,dataframe] >= 2022.01.0, <= 2024.8.0
     distributed >= 2022.01.0, <= 2024.8.0
-    dask-ml
     scipy
     zarr >= 2.10.0, != 2.11.0, != 2.11.1, != 2.11.2, < 3
     numba

sgkit/stats/pca.py

@@ -219,18 +219,18 @@ def pca(
         Principal axes defined as eigenvectors for sample covariance matrix.
         In the context of SVD, these are equivalent to the right singular
         vectors in the decomposition of a (N, M) matrix,
-        i.e. ``dask_ml.decomposition.TruncatedSVD.components_``.
+        i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.components_``.
     sample_pca_loading : [array-like, shape: (M, C)]
         Principal axes scaled by square root of eigenvalues.
         These values can also be interpreted as the correlation between the
         original variables and the unit-scaled principal axes.
     sample_pca_explained_variance : [array-like, shape: (C,)]
         Variance explained by each principal component. These values are equivalent
         to eigenvalues that result from the eigendecomposition of a (N, M) matrix,
-        i.e. ``dask_ml.decomposition.TruncatedSVD.explained_variance_``.
+        i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.explained_variance_``.
     sample_pca_explained_variance_ratio : [array-like, shape: (C,)]
         Ratio of variance explained to total variance for each principal component,
-        i.e. ``dask_ml.decomposition.TruncatedSVD.explained_variance_ratio_``.
+        i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.explained_variance_ratio_``.
 
     Examples
     --------

sgkit/variables.py

@@ -701,7 +701,7 @@ def _check_field(
         kind="f",
         __doc__="""Principal axes defined as eigenvectors for sample covariance matrix.
 In the context of SVD, these are equivalent to the right singular vectors in
-the decomposition of a (N, M) matrix., i.e. ``dask_ml.decomposition.TruncatedSVD.components_``.""",
+the decomposition of a (N, M) matrix., i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.components_``.""",
     )
 )
 
@@ -715,7 +715,7 @@ def _check_field(
         kind="f",
         __doc__="""Variance explained by each principal component. These values are equivalent
 to eigenvalues that result from the eigendecomposition of a (N, M) matrix,
-i.e. ``dask_ml.decomposition.TruncatedSVD.explained_variance_``.""",
+i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.explained_variance_``.""",
     )
 )
 
@@ -728,7 +728,7 @@ def _check_field(
         dims=("components",),
         kind="f",
         __doc__="""Ratio of variance explained to total variance for each principal component,
-i.e. ``dask_ml.decomposition.TruncatedSVD.explained_variance_ratio_``.""",
+i.e. ``sgkit.stats.truncated_svd.TruncatedSVD.explained_variance_ratio_``.""",
     )
 )