build: ruff, Dependabot and a PR labeler

.github/dependabot.yml

@@ -0,0 +1,33 @@
+# Checked monthly, with at most one grouped pull request per ecosystem.
+# For now only github-actions opens any: see the comment on the pip entry.
+version: 2
+updates:
+  - package-ecosystem: github-actions
+    directory: /
+    schedule:
+      interval: monthly
+    commit-message:
+      prefix: ci
+    groups:
+      github-actions:
+        patterns: ["*"]
+
+  # pyproject.toml and docs/requirements.txt only set lower bounds, and there is
+  # no lockfile. increase-if-necessary leaves a requirement unchanged while it
+  # allows the new release, and a lower bound allows every release, so this
+  # entry opens no version-update pull requests for now. It takes effect once a
+  # pin or an upper bound is added, and such a pull request may raise the lower
+  # bound too: tensorflow>=2.15 must stay, because TensorFlow 2.15 is how
+  # Keras 2 users run this package.
+  - package-ecosystem: pip
+    directories:
+      - /
+      - /docs
+    schedule:
+      interval: monthly
+    versioning-strategy: increase-if-necessary
+    commit-message:
+      prefix: build
+    groups:
+      python:
+        patterns: ["*"]

.github/labeler.yml

@@ -0,0 +1,63 @@
+Tests:
+- changed-files:
+  - any-glob-to-any-file: [ "tests/**/*" ]
+
+Dependencies:
+- changed-files:
+  - any-glob-to-any-file:
+    - "requirements.txt"
+    - "setup.py"
+    - "pyproject.toml"
+    - "poetry.lock"
+    - "pixi.toml"
+    - "pixi.lock"
+    - "Pipfile"
+    - "Pipfile.lock"
+    - "requirements/*.txt"
+    - "requirements/*.in"
+
+Build:
+- changed-files:
+  - any-glob-to-any-file:
+    - "Dockerfile*"
+    - "docker-compose*.yml"
+    - "Makefile"
+
+Documentation:
+- changed-files:
+  - any-glob-to-any-file:
+    - "docs/**/*"
+    - "*.md"
+    - "*.rst"
+
+Config:
+- changed-files:
+  - any-glob-to-any-file:
+    - ".pre-commit-config.yaml"
+    - "config/**/*"
+    - "settings/**/*"
+    - "*.ini"
+    - "*.cfg"
+    - "*.conf"
+
+CI/CD:
+- changed-files:
+  - any-glob-to-any-file:
+    - ".github/**/*"
+    - "tox.ini"
+    - ".coveragerc"
+
+# Add 'feature' label to any PR where the head branch name starts with `feature` or has a `feature` section in the name
+feature:
+- head-branch:
+  - '^feature'
+  - 'feature'
+
+# Add 'bugfix' label to branches with bug/fix/hotfix prefixes
+bugfix:
+- head-branch: [ '^bug', '^fix', '^hotfix' ]
+
+# Add 'breaking' label for major version bumps or breaking change commits
+breaking:
+- title: [ '^BREAKING CHANGE', 'BREAKING-CHANGE' ]
+- body: [ 'BREAKING CHANGE:', 'BREAKING-CHANGE:' ]

.github/workflows/labeler.yml

@@ -0,0 +1,12 @@
+name: "Pull Request Labeler"
+on:
+- pull_request_target
+
+jobs:
+  labeler:
+    permissions:
+      contents: read
+      pull-requests: write
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/labeler@v7.0.0

.github/workflows/test_python.yml

@@ -13,8 +13,18 @@ jobs:
       - uses: astral-sh/setup-uv@v10.3.0
         with:
           python-version: "3.12"
-      # Stop on syntax errors and undefined names only.
-      - run: uvx flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
+      # Run the ruff release that .pre-commit-config.yaml pins, so this job and the
+      # pre-commit hooks agree when pre-commit.ci updates the pin.
+      - name: Find the ruff version
+        id: ruff
+        run: |
+          version=$(grep -A1 "astral-sh/ruff-pre-commit" .pre-commit-config.yaml | sed -n "s/^ *rev: *v//p")
+          test -n "$version"
+          echo "version=$version" >> "$GITHUB_OUTPUT"
+      - name: ruff check
+        run: uvx "ruff@${{ steps.ruff.outputs.version }}" check --output-format=github .
+      - name: ruff format
+        run: uvx "ruff@${{ steps.ruff.outputs.version }}" format --check --diff .
 
   # 0.2.3 shipped invalid metadata that pip 24.1 and later refuse to install.
   # Build the package and install it the way users do before any release.

.pre-commit-config.yaml

@@ -1,4 +1,5 @@
-# Read-only checks for now; formatting and linting hooks come with the tooling update.
+# Read-only checks first, then ruff lints (fixing what it safely can) and formats
+# the Python files with the settings in pyproject.toml.
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
     rev: v6.0.0
@@ -7,3 +8,12 @@ repos:
       - id: check-toml
       - id: check-merge-conflict
       - id: check-added-large-files
+  # The lint job in .github/workflows/test_python.yml reads this rev, so keep it
+  # on the line directly below the repo line.
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.17.0
+    hooks:
+      # Lint before formatting, as ruff recommends when fixes are on.
+      - id: ruff-check
+        args: [--fix]
+      - id: ruff-format

README.md

@@ -83,7 +83,9 @@ To turn a complex NumPy array into this format:
 ```python
 import numpy as np
 
-z = np.random.randn(8, 3, 3, 2) + 1j * np.random.randn(8, 3, 3, 2)  # two complex channels
+z = np.random.randn(8, 3, 3, 2) + 1j * np.random.randn(
+    8, 3, 3, 2
+)  # two complex channels
 x = np.concatenate([z.real, z.imag], axis=-1)  # shape (8, 3, 3, 4)
 ```
 

complexnn/__init__.py

@@ -1,9 +1,6 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
-from . import bn, conv, dense, init, norm, pool
-
-# from . import fft
+from . import bn, conv, conv_utils, dense, fft, init, norm, pool, utils
 from .bn import ComplexBatchNormalization as ComplexBN
 from .conv import (
     ComplexConv,
@@ -31,3 +28,36 @@
     get_realpart,
     getpart_output_shape,
 )
+
+__all__ = [
+    "bn",
+    "conv",
+    "conv_utils",
+    "dense",
+    "fft",
+    "init",
+    "norm",
+    "pool",
+    "utils",
+    "ComplexBN",
+    "ComplexConv",
+    "ComplexConv1D",
+    "ComplexConv2D",
+    "ComplexConv3D",
+    "WeightNorm_Conv",
+    "ComplexDense",
+    "ComplexIndependentFilters",
+    "ComplexInit",
+    "IndependentFilters",
+    "SqrtInit",
+    "ComplexLayerNorm",
+    "LayerNormalization",
+    "SpectralPooling1D",
+    "SpectralPooling2D",
+    "GetAbs",
+    "GetImag",
+    "GetReal",
+    "get_imagpart",
+    "get_realpart",
+    "getpart_output_shape",
+]

complexnn/bn.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
 # Note: The implementation of complex Batchnorm is based on
 #       the Keras implementation of batch Normalization
@@ -12,11 +11,7 @@
 from tensorflow.keras.layers import InputSpec, Layer
 from tensorflow.keras.utils import register_keras_serializable
 
-from .utils import get_initializer
-
-from .utils import _ignore_input_shape_warning
-
-from .utils import _check_complex_axis
+from .utils import _check_complex_axis, _ignore_input_shape_warning, get_initializer
 
 
 def sqrt_init(shape, dtype=None):
@@ -336,7 +331,7 @@ def __init__(
         **kwargs,
     ):
         with _ignore_input_shape_warning(kwargs):
-            super(ComplexBatchNormalization, self).__init__(**kwargs)
+            super().__init__(**kwargs)
         self.supports_masking = True
         self.axis = axis
         self.momentum = momentum
@@ -359,8 +354,6 @@ def __init__(
         self.gamma_off_constraint = constraints.get(gamma_off_constraint)
 
     def build(self, input_shape):
-        ndim = len(input_shape)
-
         dim = input_shape[self.axis]
         if dim is None:
             raise ValueError(
@@ -583,5 +576,5 @@ def get_config(self):
             "gamma_diag_constraint": constraints.serialize(self.gamma_diag_constraint),
             "gamma_off_constraint": constraints.serialize(self.gamma_off_constraint),
         }
-        base_config = super(ComplexBatchNormalization, self).get_config()
+        base_config = super().get_config()
         return dict(list(base_config.items()) + list(config.items()))

complexnn/conv.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
 import numpy as np
 import tensorflow as tf
@@ -155,7 +154,7 @@ def conv_transpose_output_length(
 ):
     """Rearrange arguments for compatibility with conv_output_length."""
     if dilation != 1:
-        msg = f"Dilation must be 1 for transposed convolution. "
+        msg = "Dilation must be 1 for transposed convolution. "
         msg += f"Got dilation = {dilation}"
         raise ValueError(msg)
     # return conv_utils.deconv_length(
@@ -298,7 +297,7 @@ def __init__(
         **kwargs,
     ):
         with _ignore_input_shape_warning(kwargs):
-            super(ComplexConv, self).__init__(**kwargs)
+            super().__init__(**kwargs)
         self.rank = rank
         self.filters = filters
         self.kernel_size = conv_utils.normalize_tuple(kernel_size, rank, "kernel_size")
@@ -647,7 +646,7 @@ def get_config(self):
             "transposed": self.transposed,
             "epsilon": self.epsilon,
         }
-        base_config = super(ComplexConv, self).get_config()
+        base_config = super().get_config()
         return dict(list(base_config.items()) + list(config.items()))
 
 
@@ -760,7 +759,7 @@ def __init__(
         transposed=False,
         **kwargs,
     ):
-        super(ComplexConv1D, self).__init__(
+        super().__init__(
             rank=1,
             filters=filters,
             kernel_size=kernel_size,
@@ -784,7 +783,7 @@ def __init__(
         )
 
     def get_config(self):
-        config = super(ComplexConv1D, self).get_config()
+        config = super().get_config()
         config.pop("rank")
         # __init__ always passes data_format="channels_last" to ComplexConv.
         config.pop("data_format")
@@ -916,7 +915,7 @@ def __init__(
         transposed=False,
         **kwargs,
     ):
-        super(ComplexConv2D, self).__init__(
+        super().__init__(
             rank=2,
             filters=filters,
             kernel_size=kernel_size,
@@ -940,7 +939,7 @@ def __init__(
         )
 
     def get_config(self):
-        config = super(ComplexConv2D, self).get_config()
+        config = super().get_config()
         config.pop("rank")
         return config
 
@@ -1073,7 +1072,7 @@ def __init__(
         transposed=False,
         **kwargs,
     ):
-        super(ComplexConv3D, self).__init__(
+        super().__init__(
             rank=3,
             filters=filters,
             kernel_size=kernel_size,
@@ -1097,7 +1096,7 @@ def __init__(
         )
 
     def get_config(self):
-        config = super(ComplexConv3D, self).get_config()
+        config = super().get_config()
         config.pop("rank")
         return config
 
@@ -1122,7 +1121,7 @@ def __init__(
         epsilon=1e-07,
         **kwargs,
     ):
-        super(WeightNorm_Conv, self).__init__(**kwargs)
+        super().__init__(**kwargs)
         if self.rank == 1:
             self.data_format = "channels_last"
         self.gamma_initializer = sanitizedInitGet(gamma_initializer)
@@ -1132,7 +1131,7 @@ def __init__(
         self.gamma = None
 
     def build(self, input_shape):
-        super(WeightNorm_Conv, self).build(input_shape)
+        super().build(input_shape)
         if self.data_format == "channels_first":
             channel_axis = 1
         else:
@@ -1201,7 +1200,7 @@ def get_config(self):
             "gamma_constraint": constraints.serialize(self.gamma_constraint),
             "epsilon": self.epsilon,
         }
-        base_config = super(WeightNorm_Conv, self).get_config()
+        base_config = super().get_config()
         return dict(list(base_config.items()) + list(config.items()))
 
 

complexnn/conv_utils.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
 # Copyright 2015 The TensorFlow Authors. All Rights Reserved.
 #