Only really import packages when needed (#410)

* Only really import packages when needed

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

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

---------

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

Author: basnijholt

adaptive/learner/triangulation.py

@@ -16,9 +16,9 @@
     ones,
     square,
     subtract,
+    zeros,
 )
 from numpy import sum as np_sum
-from numpy import zeros
 from numpy.linalg import det as ndet
 from numpy.linalg import matrix_rank, norm, slogdet, solve
 

adaptive/runner.py

@@ -14,6 +14,7 @@
 import warnings
 from contextlib import suppress
 from datetime import datetime, timedelta
+from importlib.util import find_spec
 from typing import TYPE_CHECKING, Any, Callable, Union
 
 import loky
@@ -49,35 +50,32 @@
 except ImportError:
     from typing_extensions import Literal
 
-try:
-    import ipyparallel
-    from ipyparallel.client.asyncresult import AsyncResult
 
-    with_ipyparallel = True
-    ExecutorTypes: TypeAlias = Union[
-        ExecutorTypes, ipyparallel.Client, ipyparallel.client.view.ViewExecutor
-    ]
-    FutureTypes: TypeAlias = Union[FutureTypes, AsyncResult]
-except ModuleNotFoundError:
-    with_ipyparallel = False
+with_ipyparallel = find_spec("ipyparallel") is not None
+with_distributed = find_spec("distributed") is not None
+with_mpi4py = find_spec("mpi4py") is not None
 
-try:
-    import distributed
+if TYPE_CHECKING:
+    if with_distributed:
+        import distributed
 
-    with_distributed = True
-    ExecutorTypes: TypeAlias = Union[
-        ExecutorTypes, distributed.Client, distributed.cfexecutor.ClientExecutor
-    ]
-except ModuleNotFoundError:
-    with_distributed = False
+        ExecutorTypes: TypeAlias = Union[
+            ExecutorTypes, distributed.Client, distributed.cfexecutor.ClientExecutor
+        ]
 
-try:
-    import mpi4py.futures
+    if with_mpi4py:
+        import mpi4py.futures
 
-    with_mpi4py = True
-    ExecutorTypes: TypeAlias = Union[ExecutorTypes, mpi4py.futures.MPIPoolExecutor]
-except ModuleNotFoundError:
-    with_mpi4py = False
+        ExecutorTypes: TypeAlias = Union[ExecutorTypes, mpi4py.futures.MPIPoolExecutor]
+
+    if with_ipyparallel:
+        import ipyparallel
+        from ipyparallel.client.asyncresult import AsyncResult
+
+        ExecutorTypes: TypeAlias = Union[
+            ExecutorTypes, ipyparallel.Client, ipyparallel.client.view.ViewExecutor
+        ]
+        FutureTypes: TypeAlias = Union[FutureTypes, AsyncResult]
 
 with suppress(ModuleNotFoundError):
     import uvloop
@@ -934,9 +932,12 @@ def replay_log(
 
 
 def _ensure_executor(executor: ExecutorTypes | None) -> concurrent.Executor:
+    if with_ipyparallel:
+        import ipyparallel
+    if with_distributed:
+        import distributed
     if executor is None:
         executor = _default_executor()
-
     if isinstance(executor, concurrent.Executor):
         return executor
     elif with_ipyparallel and isinstance(executor, ipyparallel.Client):
@@ -955,6 +956,12 @@ def _get_ncores(
     ex: (ExecutorTypes),
 ) -> int:
     """Return the maximum  number of cores that an executor can use."""
+    if with_ipyparallel:
+        import ipyparallel
+    if with_distributed:
+        import distributed
+    if with_mpi4py:
+        import mpi4py.futures
     if with_ipyparallel and isinstance(ex, ipyparallel.client.view.ViewExecutor):
         return len(ex.view)
     elif isinstance(

docs/logo.py

@@ -3,8 +3,8 @@
 
 import holoviews
 import matplotlib.pyplot as plt
-import numpy as np
 import matplotlib.tri as mtri
+import numpy as np
 from PIL import Image, ImageDraw
 
 sys.path.insert(0, os.path.abspath(".."))  # to get adaptive on the path

docs/source/algorithms_and_examples.md

@@ -46,11 +46,13 @@ Click on the *Play* {fa}`play` button or move the sliders.
 :tags: [hide-cell]
 
 import itertools
-import adaptive
-from adaptive.learner.learner1D import uniform_loss, default_loss
+
 import holoviews as hv
 import numpy as np
 
+import adaptive
+from adaptive.learner.learner1D import default_loss, uniform_loss
+
 adaptive.notebook_extension()
 hv.output(holomap="scrubber")
 ```

docs/source/tutorial/tutorial.AverageLearner1D.md

@@ -1,14 +1,15 @@
 ---
-kernelspec:
-  name: python3
-  display_name: python3
 jupytext:
   text_representation:
     extension: .md
     format_name: myst
-    format_version: '0.13'
-    jupytext_version: 1.13.8
+    format_version: 0.13
+    jupytext_version: 1.14.5
+kernelspec:
+  display_name: python3
+  name: python3
 ---
+
 # Tutorial {class}`~adaptive.AverageLearner1D`
 
 ```{note}
@@ -23,9 +24,10 @@ import adaptive
 
 adaptive.notebook_extension()
 
+from functools import partial
+
 import holoviews as hv
 import numpy as np
-from functools import partial
 ```
 
 ## General use

docs/source/tutorial/tutorial.BalancingLearner.md

@@ -24,10 +24,11 @@ import adaptive
 
 adaptive.notebook_extension()
 
+import random
+from functools import partial
+
 import holoviews as hv
 import numpy as np
-from functools import partial
-import random
 ```
 
 The balancing learner is a “meta-learner” that takes a list of learners.

docs/source/tutorial/tutorial.Learner1D.md

@@ -25,9 +25,10 @@ import adaptive
 
 adaptive.notebook_extension()
 
-import numpy as np
-from functools import partial
 import random
+from functools import partial
+
+import numpy as np
 ```
 
 ## scalar output: `f:ℝ → ℝ`
@@ -41,8 +42,8 @@ offset = random.uniform(-0.5, 0.5)
 
 
 def f(x, offset=offset, wait=True):
-    from time import sleep
     from random import random
+    from time import sleep
 
     a = 0.01
     if wait:
@@ -155,8 +156,8 @@ To do this, you need to tell the learner to look at the curvature by specifying
 ```{code-cell} ipython3
 from adaptive.learner.learner1D import (
     curvature_loss_function,
-    uniform_loss,
     default_loss,
+    uniform_loss,
 )
 
 curvature_loss = curvature_loss_function()

docs/source/tutorial/tutorial.Learner2D.md

@@ -20,11 +20,12 @@ Download the notebook in order to see the real behaviour. [^download]
 ```{code-cell} ipython3
 :tags: [hide-cell]
 
-import adaptive
+from functools import partial
+
 import holoviews as hv
 import numpy as np
 
-from functools import partial
+import adaptive
 
 adaptive.notebook_extension()
 ```
@@ -33,9 +34,10 @@ Besides 1D functions, we can also learn 2D functions: $f: ℝ^2 → ℝ$.
 
 ```{code-cell} ipython3
 def ring(xy, wait=True):
-    import numpy as np
-    from time import sleep
     from random import random
+    from time import sleep
+
+    import numpy as np
 
     if wait:
         sleep(random() / 10)

docs/source/tutorial/tutorial.custom_loss.md

@@ -25,8 +25,8 @@ import adaptive
 adaptive.notebook_extension()
 
 # Import modules that are used in multiple cells
-import numpy as np
 import holoviews as hv
+import numpy as np
 ```
 
 {class}`~adaptive.Learner1D` and {class}`~adaptive.Learner2D` both work on the principle of subdividing their domain into subdomains, and assigning a property to each subdomain, which we call the *loss*.
@@ -137,7 +137,7 @@ def resolution_loss_function(min_distance=0, max_distance=1):
     because the total area is normalized to 1."""
 
     def resolution_loss(ip):
-        from adaptive.learner.learner2D import default_loss, areas
+        from adaptive.learner.learner2D import areas, default_loss
 
         loss = default_loss(ip)
 

pyproject.toml

@@ -98,7 +98,7 @@ python_version = "3.7"
 [tool.ruff]
 line-length = 150
 target-version = "py37"
-select = ["B", "C", "E", "F", "W", "T", "B9"]
+select = ["B", "C", "E", "F", "W", "T", "B9", "I"]
 ignore = [
     "T20",     # flake8-print
     "ANN101",  # Missing type annotation for {name} in method