📝 Fix whitespace errors on :return statements

InnerEye/Azure/azure_runner.py

@@ -107,7 +107,7 @@ def create_dataset_configs(azure_config: AzureConfig,
  :param all_dataset_mountpoints: When using the datasets in AzureML, these are the per-dataset mount points.
  :param all_local_datasets: The paths for all local versions of the datasets.
  :return: A list of DatasetConfig objects, in the same order as datasets were provided in all_azure_dataset_ids,
- omitting datasets with an empty name.
+  omitting datasets with an empty name.
  """
  datasets: List[DatasetConfig] = []
  num_local = len(all_local_datasets)

InnerEye/Azure/run_pytest.py

@@ -23,7 +23,7 @@ def run_pytest(pytest_mark: str, outputs_folder: Path) -> Tuple[bool, Path]:
  :param pytest_mark: The PyTest mark to use for filtering out the tests to run.
  :param outputs_folder: The folder into which the test result XML file should be written.
  :return: True if PyTest found tests to execute and completed successfully, False otherwise.
- Also returns the path to the generated PyTest results file.
+  Also returns the path to the generated PyTest results file.
  """
  from _pytest.main import ExitCode
  _outputs_file = outputs_folder / PYTEST_RESULTS_FILE

InnerEye/Azure/secrets_handling.py

@@ -107,7 +107,7 @@ def read_all_settings(project_settings_file: Optional[Path] = None,
  :param project_settings_file: The first YAML settings file to read.
  :param project_root: The folder that can contain a 'InnerEyePrivateSettings.yml' file.
  :return: A dictionary mapping from string to variable value. The dictionary key is the union of variable names
- found in the two settings files.
+  found in the two settings files.
  """
  private_settings_file = None
  if project_root and project_root.is_dir():
@@ -126,7 +126,7 @@ def read_settings_and_merge(project_settings_file: Optional[Path] = None,
  :param project_settings_file: The first YAML settings file to read.
  :param private_settings_file: The second YAML settings file to read. Settings in this file has higher priority.
  :return: A dictionary mapping from string to variable value. The dictionary key is the union of variable names
- found in the two settings files.
+  found in the two settings files.
  """
  result = dict()
  if project_settings_file:

InnerEye/Common/Statistics/mann_whitney_test.py

@@ -187,7 +187,7 @@ def roc_value(lst1: List[float], lst2: List[float]) -> float:
  :param lst1: a list of numbers
  :param lst2: another list of numbers
  :return: the proportion of pairs (x, y), where x is from lst1 and y is from lst2, for which
- x < y, with x == y counting as half an instance.
+  x < y, with x == y counting as half an instance.
  """
  if len(lst1) == 0 or len(lst2) == 0:
  return 0.5
@@ -273,7 +273,7 @@ def compare_scores_across_institutions(metrics_file: str, splits_to_use: str = "
  :param splits_to_use: a comma-separated list of split names
  :param mode_to_use: test, validation etc
  :return: a list of comparison lines between pairs of splits. If splits_to_use is non empty,
- only pairs involving at least one split from that set are compared.
+  only pairs involving at least one split from that set are compared.
  """
  valid_splits = set(splits_to_use.split(",")) if splits_to_use else None
  metrics = pd.read_csv(metrics_file)
@@ -330,7 +330,7 @@ def get_arguments(arglist: List[str] = None) -> Tuple[Optional[argparse.Namespac
  The value of the "-a" switch is one or more split names; pairs of splits not including these
  will not be compared.
  :return: parsed arguments and identifier for pattern (1, 2, 3 as above), or None, None if none of the
- patterns are followed
+  patterns are followed
  """
  # Use argparse because we want to have mandatory non-switch arguments, which GenericConfig doesn't support.
  parser = argparse.ArgumentParser("Run Mann-Whitney tests")

InnerEye/Common/Statistics/report_structure_extremes.py

@@ -175,7 +175,7 @@ def extent_list(presence: np.array, max_value: int) -> Tuple[List[int], List[str
  :param presence: a 1-D array of distinct integers in increasing order.
  :param max_value: any integer, not necessarily related to presence
  :return: two tuples: (1) a list of the minimum and maximum values of presence, and max_value;
- (2) a list of strings, each denoting a missing range of values within "presence".
+  (2) a list of strings, each denoting a missing range of values within "presence".
  """
  if len(presence) == 0:
  return [-1, -1, max_value], []

InnerEye/Common/Statistics/wilcoxon_signed_rank_test.py

@@ -136,7 +136,7 @@ def difference_counts(values1: List[float], values2: List[float]) -> Tuple[int,
  :param values1: list of values
  :param values2: list of values, same length as values1
  :return: number of pairs in which first value is greater than second, and number of pairs
- in which second is greater than first
+  in which second is greater than first
  """
  n1 = 0
  n2 = 0

InnerEye/Common/spawn_subprocess.py

@@ -20,7 +20,7 @@ def spawn_and_monitor_subprocess(process: str,
  :param env: The environment variables that the new process will run with. If not provided, copy the
  environment from the current process.
  :return: Return code after the process has finished, and the list of lines that were written to stdout by the
- subprocess.
+  subprocess.
  """
  if env is None:
  env = dict(os.environ.items())

InnerEye/ML/SSL/lightning_containers/ssl_container.py

@@ -204,9 +204,9 @@ def _create_ssl_data_modules(self, is_ssl_encoder_module: bool) -> InnerEyeVisio
 
  :param is_ssl_encoder_module: whether to return the data module for SSL training or for linear head. If true,
  :return transforms with two views per sample (batch like (img_v1, img_v2, label)). If False, return only one
- view per sample but also return the index of the sample in the dataset (to make sure we don't use twice the same
- batch in one training epoch (batch like (index, img_v1, label), as classifier dataloader expected to be shorter
- than SSL training, hence CombinedDataloader might loop over data several times per epoch).
+  view per sample but also return the index of the sample in the dataset (to make sure we don't use twice the same
+  batch in one training epoch (batch like (index, img_v1, label), as classifier dataloader expected to be shorter
+  than SSL training, hence CombinedDataloader might loop over data several times per epoch).
  """
  datamodule_args = self.datamodule_args[SSLDataModuleType.ENCODER] if is_ssl_encoder_module else \
  self.datamodule_args[SSLDataModuleType.LINEAR_HEAD]

InnerEye/ML/baselines_util.py

@@ -116,8 +116,8 @@ def download_and_compare_scores(outputs_folder: Path, azure_config: AzureConfig,
  :param model_dataset_df: dataframe containing contents of dataset.csv for the current model
  :param model_metrics_df: dataframe containing contents of metrics.csv for the current model
  :return: a dataframe for all the data (current model and all baselines); whether any comparisons were
- done, i.e. whether a valid baseline was found; and the text lines to be written to the Wilcoxon results
- file.
+  done, i.e. whether a valid baseline was found; and the text lines to be written to the Wilcoxon results
+  file.
  """
  comparison_baselines = get_comparison_baselines(outputs_folder, azure_config, comparison_blob_storage_paths)
  result = perform_score_comparisons(model_dataset_df, model_metrics_df, comparison_baselines)
@@ -261,7 +261,7 @@ def compare_folder_contents(expected_folder: Path,
  :param csv_relative_tolerance: When comparing CSV files, use this as the maximum allowed relative discrepancy.
  If 0.0, do not allow any discrepancy.
  :return: A list of human readable error messages, with message and file path. If no errors are found, the list is
- empty.
+  empty.
  """
  messages = []
  if run and is_offline_run_context(run):

InnerEye/ML/dataset/scalar_dataset.py

@@ -746,7 +746,7 @@ def load_item(self, item: ScalarDataSource) -> ScalarItem:
 
  :param item: The item to load.
  :return: A ClassificationItem instances with the loaded images, and the labels and non-image features copied
- from the argument.
+  from the argument.
  """
  sample = item.load_images(
  root_path=self.args.local_dataset,

InnerEye/ML/dataset/scalar_sample.py

@@ -135,7 +135,7 @@ def load_images(self,
  :param image_size: If given, all loaded images will be reshaped to the size given here, prior to the
  center crop.
  :return: An instance of ClassificationItem, with the same label and numerical_non_image_features fields,
- and all images loaded.
+  and all images loaded.
  """
  full_channel_files = self.get_all_image_filepaths(root_path=root_path,
  file_mapping=file_mapping)
@@ -160,7 +160,7 @@ def is_valid(self) -> bool:
  be not None, and none of the non imaging features may be NaN or infinity.
 
  :return: True if channel files is a list with not-None entries, and all non imaging features are finite
- floating point numbers.
+  floating point numbers.
  """
  return self.files_valid() and super().is_valid()
 

InnerEye/ML/deep_learning_config.py

@@ -855,10 +855,10 @@ def load_checkpoint_and_modify(self, path_to_checkpoint: Path) -> Dict[str, Any]
 
  :param path_to_checkpoint: Path to the checkpoint file.
  :return: Dictionary with model and optimizer state dicts. The dict should have at least the following keys:
- 1. Key ModelAndInfo.MODEL_STATE_DICT_KEY and value set to the model state dict.
- 2. Key ModelAndInfo.EPOCH_KEY and value set to the checkpoint epoch.
- Other (optional) entries corresponding to keys ModelAndInfo.OPTIMIZER_STATE_DICT_KEY and
- ModelAndInfo.MEAN_TEACHER_STATE_DICT_KEY are also supported.
+  1. Key ModelAndInfo.MODEL_STATE_DICT_KEY and value set to the model state dict.
+  2. Key ModelAndInfo.EPOCH_KEY and value set to the checkpoint epoch.
+  Other (optional) entries corresponding to keys ModelAndInfo.OPTIMIZER_STATE_DICT_KEY and
+  ModelAndInfo.MEAN_TEACHER_STATE_DICT_KEY are also supported.
  """
  return load_checkpoint(path_to_checkpoint=path_to_checkpoint, use_gpu=self.use_gpu)
 

InnerEye/ML/metrics.py

@@ -227,7 +227,7 @@ def compute_dice_across_patches(segmentation: torch.Tensor,
  :param allow_multiple_classes_for_each_pixel: If set to False, ground-truth tensor has
  to contain only one foreground label for each pixel.
  :return A torch tensor of size (Patches, Classes) with the Dice scores. Dice scores are computed for
- all classes including the background class at index 0.
+  all classes including the background class at index 0.
  """
  check_size_matches(segmentation, ground_truth, 4, 5, [0, -3, -2, -1],
  arg1_name="segmentation", arg2_name="ground_truth")

InnerEye/ML/model_config_base.py

@@ -119,7 +119,7 @@ def create_data_loaders(self) -> Dict[ModelExecutionMode, Any]:
  """
  Creates the torch DataLoaders that supply the training and the validation set during training only.
  :return: A dictionary, with keys ModelExecutionMode.TRAIN and ModelExecutionMode.VAL, and their respective
- data loaders.
+  data loaders.
  """
  logging.info("Starting to read and parse the datasets.")
  if self._datasets_for_training is None:
@@ -185,7 +185,7 @@ def get_cross_validation_dataset_splits(self, dataset_split: DatasetSplits) -> D
 
  :param dataset_split: The full dataset, split into training, validation and test section.
  :return: The dataset split with training and validation sections shuffled according to the current
- cross validation index.
+  cross validation index.
  """
  splits = dataset_split.get_k_fold_cross_validation_splits(self.number_of_cross_validation_splits)
  return splits[self.cross_validation_split_index]

InnerEye/ML/model_training.py

@@ -212,8 +212,8 @@ def model_train(checkpoint_path: Optional[Path],
  :param container: A container object that holds the training data in PyTorch Lightning format
  and the model to train.
  :return: A tuple of [Trainer, StoringLogger]. Trainer is the Lightning Trainer object that was used for fitting
- the model. The StoringLogger object is returned when training an InnerEye built-in model, this is None when
- fitting other models.
+  the model. The StoringLogger object is returned when training an InnerEye built-in model, this is None when
+  fitting other models.
  """
  lightning_model = container.model
 

InnerEye/ML/pipelines/ensemble.py

@@ -56,7 +56,7 @@ def aggregate_results(results: Iterable[InferencePipeline.Result],
 
  :param aggregation_type: aggregation function to use to combine the results.
  :return: InferenceResult: contains a Segmentation for each of the classes and their posterior
- probabilities.
+  probabilities.
  """
  if aggregation_type != EnsembleAggregationType.Average:
  raise NotImplementedError(f"Ensembling is not implemented for aggregation type: {aggregation_type}")
@@ -86,7 +86,7 @@ def predict_whole_image(self, image_channels: np.ndarray,
  :param mask: A binary image used to ignore results outside it in format: Z x Y x X.
  :param patient_id: The identifier of the patient this image belongs to.
  :return InferenceResult: that contains Segmentation for each of the classes and their posterior
- probabilities.
+  probabilities.
  """
  logging.info(f"Ensembling inference pipelines ({self._get_pipeline_ids()}) "
  f"predictions for patient: {patient_id}, "

InnerEye/ML/plotting.py

@@ -70,7 +70,7 @@ def plot_val_dice_per_epoch(metrics: Dict[str, Any]) -> int:
 
  :param metrics:
  :return: The number of series that were plotted in the graph. Can return 0 if the metrics dictionary
- does not contain any validation Dice score.
+  does not contain any validation Dice score.
  """
  plt.clf()
  series_count = 0

InnerEye/ML/reports/classification_report.py

@@ -219,8 +219,8 @@ def plot_scores_and_summary(all_labels_and_model_outputs: Sequence[LabelsAndPred
 
  :param ax: Axes object onto which to plot (default: use current axes).
  :return: A tuple of `(line_handles, summary_handle)` to use in setting a legend for the plot: `line_handles` is a
- list corresponding to the curves for each `LabelsAndPredictions`, and `summary_handle` references the median line
- and shaded CI area.
+  list corresponding to the curves for each `LabelsAndPredictions`, and `summary_handle` references the median line
+  and shaded CI area.
  """
  if ax is None:
  ax = plt.gca()

InnerEye/ML/run_ml.py

@@ -530,7 +530,7 @@ def register_model(self,
  :param checkpoint_paths: Checkpoint paths to register.
  :param model_proc: whether it's a single or ensemble model.
  :returns Tuple element 1: AML model object, or None if no model could be registered.
- Tuple element 2: The result of running the model_deployment_hook, or None if no hook was supplied.
+  Tuple element 2: The result of running the model_deployment_hook, or None if no hook was supplied.
  """
  if self.is_offline_run:
  raise ValueError("Cannot register models when InnerEye is running outside of AzureML.")
@@ -731,7 +731,7 @@ def are_sibling_runs_finished(self) -> bool:
  or cancelled.
 
  :return: True if all sibling runs of the current run have finished (they either completed successfully,
- or failed). False if any of them is still pending (running or queued).
+  or failed). False if any of them is still pending (running or queued).
  """
  if (not self.is_offline_run) \
  and (azure_util.is_cross_validation_child_run(RUN_CONTEXT)):

InnerEye/ML/runner.py

@@ -205,7 +205,7 @@ def run(self) -> Tuple[Optional[DeepLearningConfig], AzureRunInfo]:
  The main entry point for training and testing models from the commandline. This chooses a model to train
  via a commandline argument, runs training or testing, and writes all required info to disk and logs.
  :return: If submitting to AzureML, returns the model configuration that was used for training,
- including commandline overrides applied (if any).
+  including commandline overrides applied (if any).
  """
  # Usually, when we set logging to DEBUG, we want diagnostics about the model
  # build itself, but not the tons of debug information that AzureML submissions create.
@@ -450,7 +450,7 @@ def run(project_root: Path,
  The main entry point for training and testing models from the commandline. This chooses a model to train
  via a commandline argument, runs training or testing, and writes all required info to disk and logs.
  :return: If submitting to AzureML, returns the model configuration that was used for training,
- including commandline overrides applied (if any). For details on the arguments, see the constructor of Runner.
+  including commandline overrides applied (if any). For details on the arguments, see the constructor of Runner.
  """
  runner = Runner(project_root, yaml_config_file, post_cross_validation_hook, model_deployment_hook)
  return runner.run()

InnerEye/ML/utils/device_aware_module.py

@@ -25,7 +25,7 @@ def __init__(self) -> None:
  def get_devices(self) -> List[torch.device]:
  """
  :return a list of device ids on which this module
- is deployed.
+  is deployed.
  """
  return list({x.device for x in self.parameters()})
 

InnerEye/ML/utils/features_util.py

@@ -32,7 +32,7 @@ def from_data_sources(sources: List[ScalarDataSource]) -> FeatureStatistics:
 
  :param sources: list of data sources
  :return: a Feature Statistics object storing mean and standard deviation for each non-imaging feature of
- the dataset.
+  the dataset.
  """
  if len(sources) == 0:
  raise ValueError("sources must have a length greater than 0")

InnerEye/ML/utils/hdf5_util.py

@@ -75,7 +75,7 @@ def parse_acquisition_date(date: str) -> Optional[datetime]:
 
  :param date: string representing a date
  :return: converted date, None if the string is invalid for
- date conversion.
+  date conversion.
  """
  try:
  return datetime.strptime(date, DATE_FORMAT)

InnerEye/ML/utils/image_util.py

@@ -212,7 +212,7 @@ def posteriors_to_segmentation(posteriors: NumpyOrTorch) -> NumpyOrTorch:
  :param posteriors: Confidence maps [0,1] for each patch per class in format: Batches x Class x Z x Y x X
  or Class x Z x Y x X for non-batched input
  :returns segmentation: argmaxed posteriors with each voxel belonging to a single class: Batches x Z x Y x X
- or Z x Y x X for non-batched input
+  or Z x Y x X for non-batched input
  """
 
  if posteriors is None:
@@ -507,7 +507,7 @@ def compute_uncertainty_map_from_posteriors(posteriors: np.ndarray) -> np.ndarra
  :param posteriors: Normalized probability distribution in range [0, 1] for each class,
  in shape: Class x Z x Y x X
  :return: Shannon Entropy for each voxel, shape: Z x Y x X expected range is [0,1] where 1 represents
- low confidence or uniform posterior distribution across classes.
+  low confidence or uniform posterior distribution across classes.
  """
  check_if_posterior_array(posteriors)
 
@@ -569,7 +569,7 @@ def segmentation_to_one_hot(segmentation: torch.Tensor,
 
  :param result_dtype: The torch data type that the result tensor should have. This would be either float16 or float32
  :return: A torch tensor with one-hot encoding of the segmentation of shape
- [B, C*HDF5_NUM_SEGMENTATION_CLASSES, Z, Y, X]
+  [B, C*HDF5_NUM_SEGMENTATION_CLASSES, Z, Y, X]
  """
 
  def to_cuda(x: torch.Tensor) -> torch.Tensor:

InnerEye/ML/utils/io_util.py

@@ -345,7 +345,7 @@ def load_images_and_stack(files: Iterable[Path],
 
  :param image_size: If supplied, all loaded images will be resized immediately after loading.
  :return: A wrapper class that contains the loaded images, and if load_segmentation is True, also the segmentations
- that were present in the files.
+  that were present in the files.
  """
  images = []
  segmentations = []