prpy 0.2.2__py3-none-any.whl

prpy/tensorflow/signal.py

@@ -0,0 +1,103 @@
+# Copyright (c) 2024 Philipp Rouast
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in all
+# copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+# SOFTWARE.
+
+import tensorflow as tf
+from typing import Union
+
+def normalize(
+    x: Union[tf.Tensor, list],
+    axis: Union[int, tuple, None] = -1
+  ) -> tf.Tensor:
+  """Perform normalization
+
+  Args:
+    x: The input data
+    axis: Axis over which to normalize
+  Returns:
+    x: The normalized data
+  """
+  assert isinstance(x, (tf.Tensor, list))
+  assert axis is None or isinstance(axis, (int, tuple))
+  # Convert to tf.Tensor if necessary
+  if not tf.is_tensor(x):
+    x = tf.convert_to_tensor(x)
+  mean = tf.math.reduce_mean(x, axis=axis, keepdims=True)
+  return x - mean
+
+def scale(
+    x: Union[tf.Tensor, list],
+    axis: Union[int, tuple, None] = -1
+  ) -> tf.Tensor:
+  """Perform scaling
+
+  Args:
+    x: The input data
+    axis: Axis over which to scale
+  Returns:
+    x: The scaled data
+  """
+  assert isinstance(x, (tf.Tensor, list))
+  assert axis is None or isinstance(axis, (int, tuple))
+  # Convert to tf.Tensor if necessary
+  if not tf.is_tensor(x):
+    x = tf.convert_to_tensor(x)
+  std = tf.math.reduce_std(x, axis=axis, keepdims=True)
+  return tf.math.divide_no_nan(x, std)
+
+def standardize(
+    x: Union[tf.Tensor, list],
+    axis: Union[int, tuple, None] = -1
+  ) -> tf.Tensor:
+  """Perform standardization
+
+  Args:
+    x: The input data
+    axis: Axis over which to standardize
+  Returns:
+    x: The standardized data
+  """
+  assert isinstance(x, (tf.Tensor, list))
+  assert axis is None or isinstance(axis, (int, tuple))
+  # Convert to tf.Tensor if necessary
+  if not tf.is_tensor(x):
+    x = tf.convert_to_tensor(x)
+  mean = tf.math.reduce_mean(x, axis=axis, keepdims=True)
+  std = tf.math.reduce_std(x, axis=axis, keepdims=True)
+  return tf.math.divide_no_nan(x - mean, std)
+
+def diff(
+    x: Union[tf.Tensor, list],
+    axis: int = 0):
+  """Compute first signal difference.
+
+  Args:
+    x: The signal
+    axis: Scalar, the dimension across which to calculate diff.
+  Returns:
+    y: The diff signal
+  """
+  assert isinstance(x, (tf.Tensor, list))
+  # Convert to tf.Tensor if necessary
+  if not tf.is_tensor(x):
+    x = tf.convert_to_tensor(x)
+  assert axis==0 or axis==1, "Only axis=0 or axis=1 supported"
+  return tf.cond(tf.equal(axis, 0),
+    true_fn=lambda: x[1:] - x[:-1],
+    false_fn=lambda: x[:,1:] - x[:,:-1])

prpy/torch/__init__.py

@@ -0,0 +1,19 @@
+# Copyright (c) 2024 Philipp Rouast
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in all
+# copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+# SOFTWARE.

prpy/torch/model_saver.py

@@ -0,0 +1,208 @@
+# Copyright (c) 2024 Philipp Rouast
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in all
+# copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+# SOFTWARE.
+
+"""Save the best models"""
+
+import glob
+import logging
+import os
+import torch
+from typing import Union, Callable, Tuple
+
+class Candidate(object):
+  """A candidate model with a score"""
+  def __init__(
+      self,
+      score: Union[float, int],
+      dir: str,
+      filename: str
+    ):
+    """Initialise a new candidate.
+
+    Args:
+      score: The score achieved by the candidate
+      dir: The directory where the candidate model can be saved
+      filename: The filename under which the candidate model can be saved
+    """
+    assert isinstance(score, (float, int))
+    assert isinstance(dir, str)
+    assert isinstance(filename, str)
+    self.score = score
+    self.filepath = os.path.join(dir, filename)
+
+class ModelSaver(object):
+  """Save the best models to disk"""
+  def __init__(
+      self,
+      dir: str = "checkpoints",
+      keep_best: int = 5,
+      keep_latest: int = 1,
+      save_optimizer: bool = False,
+      compare_fn: Callable[[float, float], bool] = lambda x,y: x.score < y.score,
+      sort_reverse: bool = False,
+      log_fn: Callable[[str], None] = logging.info
+    ):
+    """Init the ModelSaver
+    
+    Args:
+      dir: The directory where models should be saved
+      keep_best: The number of best scoring models to keep
+      keep_latest: The number of latest models to keep
+      save_optimizer: Also save optimizer state?
+      compare_fn: Function that compares two scores
+      sort_reverse: Reverse sort order?
+      log_fn: Function to write logs
+    """
+    assert isinstance(dir, str)
+    assert isinstance(keep_best, int)
+    assert isinstance(keep_latest, int)
+    assert isinstance(save_optimizer, bool)
+    assert callable(compare_fn)
+    assert isinstance(sort_reverse, bool)
+    assert callable(log_fn)
+    self.best_candidates = []
+    self.latest_candidates = []
+    # The destination directory (make if necessary)
+    self.dir = dir
+    if not os.path.exists(self.dir):
+      os.makedirs(self.dir)
+    self.keep_best = keep_best
+    self.keep_latest = keep_latest
+    self.save_optimizer = save_optimizer
+    self.compare_fn_best = compare_fn
+    self.compare_fn_latest = lambda x,y: x.score > y.score
+    self.sort_reverse = sort_reverse
+    self.log_fn = log_fn
+
+  def __save(self, info, filepath: str):
+    """Save a model to disk.
+    
+    Args:
+      info: Tuple with torch model, optimizer, epoch, scaler, and args
+      filepath: The filepath to save to
+    """
+    assert isinstance(filepath, str)
+    # Unpack info
+    model, optimizer, epoch, scaler, args = info
+    # Save model
+    if self.save_optimizer:
+      # Save full information to support resume training
+      to_save = {
+        "model": model.state_dict(),
+        "optimizer": optimizer.state_dict(),
+        "epoch": epoch,
+        "args": args
+      }
+      if scaler is not None:
+        to_save['scaler'] = scaler.state_dict()
+    else:
+      # Save only state dict
+      to_save = {"model": model.state_dict(), "args": args}      
+    torch.save(to_save, filepath + ".pt")
+
+  def get_best_score(self) -> Union[float, None]:
+    """Get the currently best score.
+    
+    Returns:
+      score: Best score or None if no models saved
+    """
+    if len(self.best_candidates) == 0:
+      return None
+    else:
+      return self.best_candidates[0].score
+
+  def save_keep(self, info, step: int, name: str):
+    """Save and keep the given model.
+
+    Args:
+      info: Tuple with torch model, optimizer, epoch, scaler, and args
+      step: The current training step
+      name: The model name
+    """
+    assert isinstance(step, int)
+    assert isinstance(name, str)
+    self.log_fn("Saving and keeping model for step {}".format(step))
+    filepath = os.path.join(self.dir, name + "_keep_" + str(step))
+    self.__save(info=info, filepath=filepath)
+
+  def save_latest(self, info, step: int, name: str):
+    """Save the given model as currently latest.
+    
+    Args:
+      info: Tuple with torch model, optimizer, epoch, scaler, and args
+      step: The current training step
+      name: The model name
+    """
+    assert isinstance(step, int)
+    assert isinstance(name, str)
+    name = name + "_latest_" + str(step)
+    # Use step as score
+    candidate = Candidate(score=step, dir=self.dir, filename=name)
+    if len(self.latest_candidates) < self.keep_latest \
+      or self.compare_fn_latest(candidate, self.latest_candidates[-1]):
+      self.log_fn("Saving latest model for step {}".format(step))
+      # Keep candidate
+      self.latest_candidates.append(candidate)
+      self.latest_candidates = sorted(
+        self.latest_candidates, key=lambda x: x.score, reverse=True)
+      # Save candidate
+      self.__save(info, filepath=candidate.filepath)
+      # Prune candidate
+      for candidate in self.latest_candidates[self.keep_latest:]:
+        for file in glob.glob(r'{}*'.format(candidate.filepath)):
+          os.remove(file)
+      self.latest_candidates = self.latest_candidates[0:self.keep_latest]
+
+  def save_best(self, info, score: float, step: int, name: str):
+    """Save the given model as a candidate for best model.
+
+    Args:
+      model: The model to be saved as candidate for best model
+      score: The score achieved by the model
+      step: The current training step
+      name: The name of the model
+    """
+    assert isinstance(score, float)
+    assert isinstance(step, int)
+    assert isinstance(name, str)
+    self.log_fn('Saving model for step {}'.format(step))
+    name = name + "_best_" + str(step)
+    candidate = Candidate(score, dir=self.dir, filename=name)
+    if len(self.best_candidates) < self.keep_best \
+      or self.compare_fn_best(candidate, self.best_candidates[-1]):
+      # Keep candidate
+      self.log_fn("Keeping model {} with score {:.4f}".format(
+        candidate.filepath, candidate.score))
+      self.best_candidates.append(candidate)
+      self.best_candidates = sorted(
+        self.best_candidates, key=lambda x: x.score, reverse=self.sort_reverse)
+      # Save candidate
+      self.__save(info, filepath=candidate.filepath)
+      # Prune candidates
+      for candidate in self.best_candidates[self.keep_best:]:
+        self.log_fn('Removing old model {} with score {:.4f}'.format(
+          candidate.filepath, candidate.score))
+        for file in glob.glob(r'{}*'.format(candidate.filepath)):
+          os.remove(file)
+      self.best_candidates = self.best_candidates[0:self.keep_best]
+    else:
+      # Skip the candidate
+      self.log_fn('Skipping candidate {}'.format(candidate.filepath))
+      

prpy-0.2.2.dist-info/LICENSE

@@ -0,0 +1,19 @@
+Copyright (c) 2024 Philipp Rouast
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

prpy-0.2.2.dist-info/METADATA

@@ -0,0 +1,75 @@
+Metadata-Version: 2.1
+Name: prpy
+Version: 0.2.2
+Summary: Collection of Python utils for signal, image, and video processing
+Author-email: Philipp Rouast <philipp@rouast.com>
+License: MIT License
+Project-URL: Repository, https://github.com/prouast/prpy.git
+Project-URL: Issues, https://github.com/prouast/prpy/issues
+Keywords: python,numpy,ffmpeg,tensorflow,torch
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Classifier: Development Status :: 4 - Beta
+Classifier: License :: OSI Approved :: MIT License
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: ffmpeg
+Requires-Dist: ffmpeg-python ; extra == 'ffmpeg'
+Requires-Dist: numpy ; extra == 'ffmpeg'
+Provides-Extra: numpy
+Requires-Dist: numpy ; extra == 'numpy'
+Requires-Dist: scipy ; extra == 'numpy'
+Requires-Dist: Pillow ; extra == 'numpy'
+Requires-Dist: opencv-python ; extra == 'numpy'
+Requires-Dist: scikit-learn ; extra == 'numpy'
+Provides-Extra: numpy_min
+Requires-Dist: numpy ; extra == 'numpy_min'
+Requires-Dist: scipy ; extra == 'numpy_min'
+Requires-Dist: Pillow ; extra == 'numpy_min'
+Provides-Extra: tensorflow
+Requires-Dist: tensorflow ==2.12.1 ; extra == 'tensorflow'
+Requires-Dist: numpy ; extra == 'tensorflow'
+Provides-Extra: test
+Requires-Dist: pytest ; extra == 'test'
+Requires-Dist: flake8 ; extra == 'test'
+Provides-Extra: torch
+Requires-Dist: torch ==2.1.0 ; extra == 'torch'
+
+[![Tests](https://github.com/prouast/prpy/actions/workflows/main.yml/badge.svg)](https://github.com/prouast/prpy/actions/workflows/main.yml)
+
+# prpy
+
+A collection of Python utilities for signal, image, and video processing.
+It contains subpackages for working with `numpy`, `ffmpeg`, `tensorflow`, and `torch`.
+
+## Installation
+
+General prerequisites are `python>=3.8` and `ffmpeg` installed and accessible via the `$PATH` environment variable.
+
+To install `prpy` and its dependencies:
+
+```
+git clone https://github.com/prouast/prpy.git
+pip install "./prpy[ffmpeg,numpy,tensorflow,torch,test]"
+```
+
+The above runs a full install of all dependencies.
+It is possible to customize the install of the dependencies by only listing the desired subpackages out of `ffmpeg`, `numpy`, `tensorflow`, `torch`, and `test` in the square brackets above.
+
+## Linting and tests
+
+To lint and run tests:
+
+```
+flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
+pytest
+```
+
+## Build
+
+To build:
+
+```
+python -m build
+```

prpy-0.2.2.dist-info/RECORD

@@ -0,0 +1,27 @@
+prpy/__init__.py,sha256=k16PtlPaiqybAKA7he03xqNVCa1v6yhIpRpjJvnTLzg,1093
+prpy/constants.py,sha256=bApkiDwFer5shNg14nEsA5g5qhRA3oFQ5Ptx2IeAWfY,1209
+prpy/ffmpeg/__init__.py,sha256=k16PtlPaiqybAKA7he03xqNVCa1v6yhIpRpjJvnTLzg,1093
+prpy/ffmpeg/probe.py,sha256=Qfxl2t4-DoaX7i3cQemiwHEwIL-DPKV4wNHnvR30gWE,3126
+prpy/ffmpeg/readwrite.py,sha256=wX3E4yaPCBDmUZGUVppDevEVf7GyPqhQw0tE33mId7g,16971
+prpy/ffmpeg/utils.py,sha256=1wBnHw9njbXMXZiQCwJx60_zKy7Kj4wAtvyvVsi6DYI,3195
+prpy/numpy/__init__.py,sha256=k16PtlPaiqybAKA7he03xqNVCa1v6yhIpRpjJvnTLzg,1093
+prpy/numpy/face.py,sha256=lCK35xHoEow-_cgMtotj7kxsD_mWNOqtSOmWpgmEddI,8314
+prpy/numpy/image.py,sha256=Sxwe8-rAmuFCxRUki5P0sQUlgcKuXdKNZvDN4nmfRXY,7064
+prpy/numpy/metric.py,sha256=nXIQaNET2-yQOslIPcMKGi3w2_XIeHkiyO1M78gS5tQ,6395
+prpy/numpy/signal.py,sha256=hkxr0paVkKvLOrqny_2Kh_pbKK4GrVBRn_EjO0ogyxU,21767
+prpy/numpy/stride_tricks.py,sha256=sDpEyIfx6R81kjtVaWL-lXVa8gQu8_-s0Ky8-QJzJ-M,7764
+prpy/tensorflow/__init__.py,sha256=k16PtlPaiqybAKA7he03xqNVCa1v6yhIpRpjJvnTLzg,1093
+prpy/tensorflow/image.py,sha256=2EM2G76C2qXj5b3ZgeA5_5oe7yO6NJJRuCyrYYFCXZ0,7554
+prpy/tensorflow/loss.py,sha256=IUI7P38IJHdvsdxd2NAOADhpN6xVNEloOZBWIDcuViA,4248
+prpy/tensorflow/lr_schedule.py,sha256=18VLjCmMICngNH0NsPC7zUekwA1X1v5WN7y_pVFvrFE,3384
+prpy/tensorflow/model_saver.py,sha256=sErMFd88gfxRcuSDoO7zWjIEvW4ZNW4MmvvjFZmWJSM,7660
+prpy/tensorflow/nan.py,sha256=4DnZxkQC7_lOdhyndyxtDqsKeyHvcDJ3n6f02gdQQzQ,8525
+prpy/tensorflow/optimizer.py,sha256=sFglHF4fCdvpjt0fAgGzye7MpqM343S4xjboRUF1zR4,9844
+prpy/tensorflow/signal.py,sha256=2F2OPNTvtf3Z5V94pk4RDLS_t0YwvhXHjYGBBcRd6a8,3326
+prpy/torch/__init__.py,sha256=k16PtlPaiqybAKA7he03xqNVCa1v6yhIpRpjJvnTLzg,1093
+prpy/torch/model_saver.py,sha256=ScN73UfVpwl7M_5uSPg8gfY2DTODxgkVR8-iuvqVZlE,7582
+prpy-0.2.2.dist-info/LICENSE,sha256=y4jYVUDrsgKYhTzvNU5lVhyZa7MEWRPg1caumfBviTc,1058
+prpy-0.2.2.dist-info/METADATA,sha256=oX4lSZK7as0wAYExBlvhHH2wR8B41ZojQsf-ICa3v-A,2465
+prpy-0.2.2.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+prpy-0.2.2.dist-info/top_level.txt,sha256=tWXjV62yfTm26Dhnt6xUpyHXt1GfB2TObPnQz8WJeg0,5
+prpy-0.2.2.dist-info/RECORD,,

prpy-0.2.2.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: bdist_wheel (0.43.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

prpy-0.2.2.dist-info/top_level.txt

@@ -0,0 +1 @@
+prpy