gnu: python-dask: Move to python-science.

* gnu/packages/python-xyz.scm (python-dask, python-dask/bootstrap): Move
from here ...
* gnu/packages/python-science.scm: ... to here.

Change-Id: I6bedc89bd09b3f2dfc3f7cbe27eb5790f32f6af7

1 files changed, 125 insertions, 0 deletions

diff --git a/gnu/packages/python-science.scm b/gnu/packages/python-science.scm
index e0ddeb6e4c..dbbd44db70 100644
--- a/gnu/packages/python-science.scm
+++ b/gnu/packages/python-science.scm
@@ -702,6 +702,131 @@ it can be used for displaying many qualitatively different samples.")
 optimization problems in Python.")
     (license license:asl2.0)))
 
+;; Note: Remember to update python-distributed when updating dask.
+(define-public python-dask
+  (package
+    (name "python-dask")
+    (version "2024.12.1")
+    (source
+     (origin
+       (method git-fetch)
+       (uri (git-reference
+             (url "https://github.com/dask/dask/")
+             (commit version)))
+       (file-name (git-file-name name version))
+       (sha256
+        (base32 "17iqfyjphyn72xdr8fmynzvixskbq16pwmsknwc6anq7s2axvas2"))))
+    (build-system pyproject-build-system)
+    (arguments
+     (list
+      ;; Avoid coverage
+      #:test-flags
+      #~(list "--numprocesses" (number->string (parallel-job-count))
+              "-m" "not gpu and not slow and not network"
+              ;; These all fail with different hashes.  Doesn't seem
+              ;; problematic.
+              "--ignore-glob=**/test_tokenize.py"
+              ;; ORC tests crash Python with a failure to find the global
+              ;; localtime file.  See also
+              ;; https://github.com/apache/arrow/issues/40633.
+              "--ignore-glob=**/test_orc.py"
+              "-k" (string-append
+                    ;; This one cannot be interrupted.
+                    "not test_interrupt"
+                    ;; This one fails with "local variable 'ctx' referenced
+                    ;; before assignment".  Maybe enable this in later
+                    ;; versions (or when pandas has been upgraded.
+                    " and not test_dt_accessor"
+                    ;; This fails when dask-expr is among the inputs.
+                    " and not test_groupby_internal_repr"
+                    ;; This fails with different job ids.
+                    " and not test_to_delayed_optimize_graph"
+                    ;; This one expects a deprecation warning that never
+                    ;; comes.
+                    " and not test_RandomState_only_funcs"
+                    ;; This test expects a RuntimeWarning that is never
+                    ;; raised.
+                    " and not test_nanquantile_all_nan")
+              ;; Tests must run from the output directory, because otherwise
+              ;; it complains about the difference between the target
+              ;; directory embedded in the pyc files and the source directory
+              ;; from which we run tests.
+              (getcwd))
+      #:phases
+      #~(modify-phases %standard-phases
+          (add-after 'unpack 'versioneer
+            (lambda _
+              ;; Our version of versioneer needs setup.cfg.  This is adapted
+              ;; from pyproject.toml.
+              (with-output-to-file "setup.cfg"
+                (lambda ()
+                  (display "\
+[versioneer]
+VCS = git
+style = pep440
+versionfile_source = dask/_version.py
+versionfile_build = dask/_version.py
+tag_prefix =
+parentdir_prefix = dask-
+")))
+              (invoke "versioneer" "install")
+              (substitute* "setup.py"
+                (("versioneer.get_version\\(\\)")
+                 (string-append "\"" #$version "\"")))))
+          (add-after 'unpack 'fix-pytest-config
+            (lambda _
+              ;; This option is not supported by our version of pytest.
+              (substitute* "pyproject.toml"
+                (("--cov-config=pyproject.toml") ""))))
+          (add-after 'unpack 'patch-pyproject
+            (lambda _
+              ;; We use pyarrow > 14
+              (substitute* "pyproject.toml"
+                (("\"pyarrow_hotfix\",") ""))))
+          (add-before 'check 'pre-check
+            (lambda _ (chdir "/tmp"))))))
+    (propagated-inputs
+     (list python-click ;needed at runtime
+           python-cloudpickle
+           python-dask-expr
+           python-fsspec
+           python-importlib-metadata ;needed at runtime for dask/_compatibility.py
+           python-numpy
+           python-packaging
+           python-pandas
+           python-partd
+           python-toolz
+           python-pyyaml))
+    (native-inputs
+     (list python-importlib-metadata
+           python-pytest
+           python-pytest-rerunfailures
+           python-pytest-runner
+           python-pytest-xdist
+           python-versioneer
+           python-wheel))
+    (home-page "https://github.com/dask/dask/")
+    (synopsis "Parallel computing with task scheduling")
+    (description
+     "Dask is a flexible parallel computing library for analytics.  It
+consists of two components: dynamic task scheduling optimized for computation,
+and large data collections like parallel arrays, dataframes, and lists that
+extend common interfaces like NumPy, Pandas, or Python iterators to
+larger-than-memory or distributed environments.  These parallel collections
+run on top of the dynamic task schedulers.")
+    (license license:bsd-3)))
+
+(define-public python-dask/bootstrap
+  (package
+    (inherit python-dask)
+    (properties '((hidden? . #true)))
+    (arguments
+     (substitute-keyword-arguments (package-arguments python-dask)
+       ((#:tests? _ #t) #f)))
+    (propagated-inputs
+     (modify-inputs (package-propagated-inputs python-dask)
+       (delete "python-dask-expr")))))
+
 (define-public python-dask-expr
   (package
     (name "python-dask-expr")