Support TF 2.12, drop Python 3.7 (adriangb#302)

.github/workflows/tests.yaml

@@ -29,11 +29,12 @@ jobs:
       - uses: pre-commit/[email protected]
 
   TestStable:
-    name: Ubuntu / Python ${{ matrix.python-version }} / TensorFlow Stable / Scikit-Learn Stable
+    name: Ubuntu / Python ${{ matrix.python-version }} / TensorFlow ${{ matrix.tensorflow-version }} / Scikit-Learn Stable
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: ["3.7", "3.8", "3.9", "3.10"]
+        python-version: ["3.8", "3.9", "3.10", "3.11"]
+        tensorflow-version: ["2.11", "2.12"]
       fail-fast: false
 
     steps:
@@ -52,7 +53,8 @@ jobs:
 
       - name: Install Dependencies
         run: |
-          poetry install -E tensorflow
+          poetry install
+          poetry run python -m pip install -U "tensorflow~=${{ matrix.tensorflow-version }}"
 
       - name: Test with pytest
         run: |
@@ -97,6 +99,7 @@ jobs:
 
       - name: Test with pytest
         if: always()
+        continue-on-error: true
         run: |
           poetry run python -m pip freeze
           poetry run python -m pytest -v --cov=scikeras --cov-report xml --color=yes
@@ -145,8 +148,9 @@ jobs:
     runs-on: ${{ matrix.os }}-latest
     strategy:
       matrix:
-        os: [MacOS, Windows]  # test all OSs (except Ubuntu, which is already running other tests)
-        python-version: ["3.7", "3.10"]  # test only the two extremes of supported Python versions
+        os: [MacOS, Windows]  # test all OSs except Ubuntu, which is already running other tests
+        python-version: ["3.8", "3.11"]  # test only the two extremes of supported Python versions
+        tensorflow-version: ["2.11", "2.12"]  # test only the two extremes of supported TF versions
       fail-fast: false
 
     steps:
@@ -166,12 +170,12 @@ jobs:
       - name: Install Dependencies
         # TF is sorta dropping support for Windows
         # At the very least they are no longer suppporting GPU
-        # See https://github.com/tensorflow/tensorflow/releases/tag/v2.11.0
+        # See https://github.com/tensorflow/tensorflow/releases/tag/v2.12.0
         # and
         # https://www.tensorflow.org/install/pip#windows-native
         run: |
           poetry install
-          poetry run pip install -U tensorflow-cpu
+          poetry run python -m pip install -U "tensorflow~=${{ matrix.tensorflow-version }}"
 
       - name: Test with pytest
         run: |

.pre-commit-config.yaml

@@ -1,12 +1,10 @@
 repos:
   - repo: https://github.com/psf/black
-    rev: 22.3.0
+    rev: "23.3.0"
     hooks:
       - id: black
 
-  - repo: https://github.com/timothycrosley/isort
-    rev: 5.10.1
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.0.272
     hooks:
-      - id: isort
-        additional_dependencies: [toml]
-        exclude: ^.*/?setup\.py$
+      - id: ruff

docs/source/conf.py

@@ -13,12 +13,11 @@
 import os
 import sys
 
-
 sys.path.insert(0, "../../scikeras")
 
 # -- Project information -----------------------------------------------------
 
-from scikeras import __version__
+from scikeras import __version__  # noqa: 402
 
 
 project = "SciKeras"
@@ -109,11 +108,9 @@ def setup(app):
 
 # Functionality to build github source URI, taken from sklearn.
 
-import inspect
-import subprocess
-
-from operator import attrgetter
-
+import inspect  # noqa: 402
+import subprocess  # noqa: 402
+from operator import attrgetter  # noqa: E402
 
 REVISION_CMD = "git rev-parse --short HEAD"
 

pyproject.toml

@@ -27,11 +27,11 @@ license = "MIT"
 name = "scikeras"
 readme = "README.md"
 repository = "https://github.com/adriangb/scikeras"
-version = "0.10.0"
+version = "0.11.0"
 
 [tool.poetry.dependencies]
 importlib-metadata = {version = ">=3", python = "<3.8"}
-python = ">=3.7.0,<3.11.0"
+python = ">=3.8.0,<3.12.0"
 scikit-learn = ">=1.0.0"
 packaging = ">=0.21"
 tensorflow = {version = ">=2.11.0", optional = true}
@@ -43,6 +43,11 @@ grpcio = { version = "<1.50.0", markers = "python_version >= '3.10' and sys_plat
 tensorflow = ["tensorflow"]
 tensorflow-cpu = ["tensorflow-cpu"]
 
+[tool.poetry.dependencies.tensorflow-io-gcs-filesystem]
+# see https://github.com/tensorflow/tensorflow/issues/60202
+version = ">=0.23.1,<0.32"
+markers = "sys_platform == 'win32'"
+
 [tool.poetry.dev-dependencies]
 tensorflow = ">=2.11.0"
 coverage = {extras = ["toml"], version = ">=6.4.2"}
@@ -58,25 +63,24 @@ pytest = ">=7.1.2"
 pytest-cov = ">=3.0.0"
 sphinx = ">=5.0.2"
 
-[tool.isort]
-atomic = true
-filter_files = true
-include_trailing_comma = true
-known_first_party = "scikeras"
-known_third_party = [
-  "tensorflow",
-  "sklearn",
+[tool.ruff]
+select = [
+    "E",  # pycodestyle errors
+    "W",  # pycodestyle warnings
+    "F",  # pyflakes
+    "I",  # isort
+    "C",  # flake8-comprehensions
+    "B",  # flake8-bugbear
+]
+ignore = [
+    "E501",  # line too long, handled by black
+    "C901",  # too complex
+    "B905",  # strict argument to zip
 ]
-line_length = 88
-lines_after_imports = 2
-lines_between_types = 1
-multi_line_output = 3
-skip_glob = ["*/setup.py"]
-use_parentheses = true
 
 [tool.black]
 line-length = 88
-target-version = ['py36', 'py38']
+target-version = ['py38']
 
 [tool.coverage.run]
 source = ["scikeras/"]

scikeras/__init__.py

@@ -13,21 +13,19 @@
 MIN_TF_VERSION = "2.7.0"
 TF_VERSION_ERR = f"SciKeras requires TensorFlow >= {MIN_TF_VERSION}."
 
-from packaging import version
-
+from packaging import version  # noqa: E402
 
 try:
     from tensorflow import __version__ as tf_version
 except ImportError:  # pragma: no cover
-    raise ImportError("TensorFlow is not installed. " + TF_VERSION_ERR)
+    raise ImportError("TensorFlow is not installed. " + TF_VERSION_ERR) from None
 else:
     if version.parse(tf_version) < version.parse(MIN_TF_VERSION):  # pragma: no cover
-        raise ImportError(TF_VERSION_ERR)
-
-import tensorflow.keras as _keras
+        raise ImportError(TF_VERSION_ERR) from None
 
-from scikeras import _saving_utils
+import tensorflow.keras as _keras  # noqa: E402
 
+from scikeras import _saving_utils  # noqa: E402
 
 _keras.Model.__reduce__ = _saving_utils.pack_keras_model
 _keras.Model.__deepcopy__ = _saving_utils.deepcopy_model

scikeras/_saving_utils.py

@@ -2,15 +2,13 @@
 import shutil
 import tarfile
 import tempfile
-
 from contextlib import contextmanager
 from io import BytesIO
-from typing import Any, Callable, ContextManager, Dict, Hashable, Iterator, List, Tuple
+from typing import Any, Callable, Dict, Hashable, Iterator, List, Tuple
 from uuid import uuid4
 
 import numpy as np
 import tensorflow.keras as keras
-
 from tensorflow import io as tf_io
 from tensorflow.keras.models import load_model
 

scikeras/_utils.py

@@ -1,13 +1,11 @@
 import inspect
-
 from types import FunctionType
 from typing import Any, Callable, Dict, Iterable, Mapping, Sequence, Type, Union
 
 from tensorflow.keras import losses as losses_mod
 from tensorflow.keras import metrics as metrics_mod
 from tensorflow.keras import optimizers as optimizers_mod
 
-
 DIGITS = frozenset(str(i) for i in range(10))
 
 
@@ -37,7 +35,7 @@ def route_params(
     Dict[str, Any]
         Filtered parameters, with any routing prefixes removed.
     """
-    res = dict()
+    res = {}
     routed = {k: v for k, v in params.items() if "__" in k}
     non_routed = {k: params[k] for k in (params.keys() - routed.keys())}
     for key, val in non_routed.items():
@@ -92,7 +90,7 @@ def unflatten_params(items, params, base_params=None):
     if inspect.isclass(items):
         item = items
         new_base_params = {p: v for p, v in params.items() if "__" not in p}
-        base_params = base_params or dict()
+        base_params = base_params or {}
         args_and_kwargs = {**base_params, **new_base_params}
         for p, v in args_and_kwargs.items():
             args_and_kwargs[p] = unflatten_params(
@@ -110,7 +108,7 @@ def unflatten_params(items, params, base_params=None):
         return item(*args, **kwargs)
     if isinstance(items, (list, tuple)):
         iter_type_ = type(items)
-        res = list()
+        res = []
         new_base_params = {p: v for p, v in params.items() if "__" not in p}
         for idx, item in enumerate(items):
             item_params = route_params(
@@ -126,7 +124,7 @@ def unflatten_params(items, params, base_params=None):
             )
         return iter_type_(res)
     if isinstance(items, (dict,)):
-        res = dict()
+        res = {}
         new_base_params = {p: v for p, v in params.items() if "__" not in p}
         for key, item in items.items():
             item_params = route_params(
@@ -144,7 +142,7 @@ def unflatten_params(items, params, base_params=None):
     # non-compilable item, check if it has any routed parameters
     item = items
     new_base_params = {p: v for p, v in params.items() if "__" not in p}
-    base_params = base_params or dict()
+    base_params = base_params or {}
     kwargs = {**base_params, **new_base_params}
     if kwargs:
         raise TypeError(

scikeras/utils/__init__.py

@@ -59,7 +59,9 @@ def loss_name(loss: Union[str, Loss, Callable]) -> str:
     fn_or_cls = keras_loss_get(loss)
     if isinstance(fn_or_cls, Loss):
         return _camel2snake(fn_or_cls.__class__.__name__)
-    return fn_or_cls.__name__
+    if hasattr(fn_or_cls, "__name__"):
+        return fn_or_cls.__name__
+    return fn_or_cls
 
 
 def metric_name(metric: Union[str, Metric, Callable]) -> str:
@@ -109,4 +111,6 @@ def metric_name(metric: Union[str, Metric, Callable]) -> str:
     fn_or_cls = keras_metric_get(metric)
     if isinstance(fn_or_cls, Metric):
         return _camel2snake(fn_or_cls.__class__.__name__)
-    return fn_or_cls.__name__
+    if hasattr(fn_or_cls, "__name__"):
+        return fn_or_cls.__name__
+    return fn_or_cls

scikeras/utils/random_state.py

@@ -1,16 +1,13 @@
 import os
 import random
-
 from contextlib import contextmanager
 from typing import Generator
 
 import numpy as np
 import tensorflow as tf
-
 from tensorflow.python.eager import context
 from tensorflow.python.framework import config, ops
 
-
 DIGITS = frozenset(str(i) for i in range(10))
 
 

scikeras/utils/transformers.py

@@ -2,7 +2,6 @@
 
 import numpy as np
 import tensorflow as tf
-
 from sklearn.base import BaseEstimator, TransformerMixin
 from sklearn.exceptions import NotFittedError
 from sklearn.pipeline import make_pipeline

scikeras/wrappers.py

@@ -3,13 +3,11 @@
 import functools
 import inspect
 import warnings
-
 from collections import defaultdict
 from typing import Any, Callable, Dict, Iterable, List, Mapping, Set, Tuple, Type, Union
 
 import numpy as np
 import tensorflow as tf
-
 from scipy.sparse import isspmatrix, lil_matrix
 from sklearn.base import BaseEstimator, ClassifierMixin, RegressorMixin
 from sklearn.exceptions import NotFittedError
@@ -219,7 +217,6 @@ def __init__(
         epochs: int = 1,
         **kwargs,
     ):
-
         # Parse hardcoded params
         self.model = model
         self.build_fn = build_fn
@@ -300,7 +297,8 @@ def _check_model_param(self):
             model = build_fn
             warnings.warn(
                 "``build_fn`` will be renamed to ``model`` in a future release,"
-                " at which point use of ``build_fn`` will raise an Error instead."
+                " at which point use of ``build_fn`` will raise an Error instead.",
+                stacklevel=4,
             )
         if model is None:
             # no model, use this class' _keras_build_fn
@@ -515,7 +513,7 @@ def _fit_keras_model(
                     except AttributeError:
                         raise ValueError(
                             f"`{bs_kwarg}=-1` requires that `X` implement `shape`"
-                        )
+                        ) from None
         fit_args = {k: v for k, v in fit_args.items() if not k.startswith("callbacks")}
         fit_args["callbacks"] = self._fit_callbacks
 
@@ -828,7 +826,6 @@ def initialize(destination: str):
     def _initialize(
         self, X: np.ndarray, y: Union[np.ndarray, None] = None
     ) -> Tuple[np.ndarray, np.ndarray]:
-
         # Handle random state
         if isinstance(self.random_state, np.random.RandomState):
             # Keras needs an integer
@@ -1016,7 +1013,7 @@ def _predict_raw(self, X, **kwargs):
                 except AttributeError:
                     raise ValueError(
                         "`batch_size=-1` requires that `X` implement `shape`"
-                    )
+                    ) from None
 
         # predict with Keras model
         y_pred = self.model_.predict(x=X, **pred_args)

tests/__init__.py

@@ -4,7 +4,6 @@
 import pytest
 import tensorflow as tf
 
-
 # Force data conversion warnings to be come errors
 pytestmark = pytest.mark.filterwarnings(
     "error::sklearn.exceptions.DataConversionWarning"

tests/mlp_models.py

@@ -3,8 +3,6 @@
 from tensorflow.keras.layers import Dense, Input
 from tensorflow.keras.models import Model
 
-from scikeras.wrappers import KerasRegressor
-
 
 def dynamic_classifier(
     hidden_layer_sizes=(10,),