pipeline.utils package

This package defines common utilities to handle the models, graphs, configurations, plots, etc.

Subpackages

• pipeline.utils.commonutils package
  • pipeline.utils.commonutils.cdetector module
    • get_coordinate_names()
  • pipeline.utils.commonutils.cfeatures module
    • get_input_features()
    • get_number_input_features()
    • get_unnormalised_features()
  • pipeline.utils.commonutils.config module
    • CommonDirs
      • CommonDirs.common_config
      • CommonDirs.detectors
      • CommonDirs.get_filenames_from_detector()
      • CommonDirs.repository
      • CommonDirs.test_config_path
    • PipelineConfig
      • PipelineConfig.add_config()
      • PipelineConfig.common_config
      • PipelineConfig.data_experiment_dir
      • PipelineConfig.detector
      • PipelineConfig.dict()
      • PipelineConfig.dir_path
      • PipelineConfig.experiment_name
      • PipelineConfig.get_test_batch_dir()
      • PipelineConfig.performance_dir
      • PipelineConfig.required_test_dataset_names
      • PipelineConfig.steps
    • get_detector_from_experiment_name()
    • get_detector_from_pipeline_config()
    • get_performance_directory_experiment()
    • get_pipeline_config_path()
    • load_config()
    • load_dict()
    • resolve_relative_path()
  • pipeline.utils.commonutils.crun module
    • InOutFunction
    • run_for_different_partitions()
  • pipeline.utils.commonutils.ctests module
    • collect_test_samples()
    • get_available_test_dataset_names()
    • get_preprocessed_test_dataset_dir()
    • get_required_test_dataset_names()
    • get_test_batch_dir()
    • get_test_batch_paths()
    • get_test_config_for_preprocessing()
    • load_preprocessing_test_config()
• pipeline.utils.graphutils package
  • pipeline.utils.graphutils.batch2df module
    • get_df_edges()
    • get_df_edges_from_edge_index()
    • get_df_hits()
    • get_df_hits_particles()
    • get_df_hits_particles_from_particle_id_hit_idx()
    • get_df_triplets_from_triplet_index()
    • merge_df_hits_particles_to_edges()
  • pipeline.utils.graphutils.edgebuilding module
    • get_random_pairs_plane_by_plane()
  • pipeline.utils.graphutils.edgeutils module
    • compute_edge_labels_from_pid_only()
    • remove_duplicate_edges()
    • sort_edge_nodes()
  • pipeline.utils.graphutils.knn module
    • build_edges_exatrkx()
    • build_edges_faiss()
    • build_edges_plane_by_plane()
  • pipeline.utils.graphutils.torchutils module
    • get_groupby_indices()
    • scatter_reduce()
  • pipeline.utils.graphutils.tripletbuilding module
    • compute_triplet_truths()
    • from_df_edges_to_df_triplets()
    • from_edge_index_to_triplet_indices()
    • from_triplet_index_to_triplet_truth()
    • get_triplet_truths_from_tensors()
  • pipeline.utils.graphutils.truths module
    • get_truth_cudf()
    • get_truths_exatrkx()
    • get_truths_pytorch()
• pipeline.utils.loaderutils package
  • pipeline.utils.loaderutils.dataiterator module
    • LazyDatasetBase
      • LazyDatasetBase.fetch_dataset()
  • pipeline.utils.loaderutils.pathandling module
    • get_input_paths()
  • pipeline.utils.loaderutils.preprocessing module
    • cast_boolean_columns()
    • combine_run_event_into_event_id()
    • load_dataframes()
    • load_preprocessed_dataframes()
  • pipeline.utils.loaderutils.tracks module
    • get_tracks_from_batch()
    • get_tracks_input_directory()
    • load_tracks()
    • load_tracks_event()
    • load_tracks_preprocessed_dataframes_given_partition()
• pipeline.utils.modelutils package
  • pipeline.utils.modelutils.basemodel module
    • ModelBase
      • ModelBase.configure_optimizers()
      • ModelBase.feature_indices
      • ModelBase.feature_means
      • ModelBase.feature_names
      • ModelBase.feature_scales
      • ModelBase.fetch_datasets()
      • ModelBase.fetch_partition()
      • ModelBase.get_features()
      • ModelBase.get_lazy_dataset()
      • ModelBase.get_lazy_dataset_partition()
      • ModelBase.get_model_from_checkpoint()
      • ModelBase.get_n_features()
      • ModelBase.get_subnetwork_inputs()
      • ModelBase.get_subnetwork_outputs()
      • ModelBase.input_kwargs
      • ModelBase.input_to_dynamic_axes
      • ModelBase.lazy
      • ModelBase.load_partition()
      • ModelBase.load_testset_from_directory()
      • ModelBase.load_trainset_split_indices()
      • ModelBase.n_trainable_params
      • ModelBase.on_step
      • ModelBase.optimizer_step()
      • ModelBase.setup()
      • ModelBase.subnetwork_groups
      • ModelBase.subnetwork_to_outputs
      • ModelBase.subnetworks
      • ModelBase.test_dataloader()
      • ModelBase.testset
      • ModelBase.to_onnx()
      • ModelBase.train_dataloader()
      • ModelBase.trainset
      • ModelBase.val_dataloader()
      • ModelBase.valset
    • ModelONNXExport
      • ModelONNXExport.model
      • ModelONNXExport.subnetwork
      • ModelONNXExport.forward()
    • check_and_discard()
    • feature_to_compute_fct
  • pipeline.utils.modelutils.batches module
    • get_batches()
    • select_subset()
  • pipeline.utils.modelutils.build module
    • BuilderBase
      • BuilderBase.build_features()
      • BuilderBase.build_weights()
      • BuilderBase.construct_downstream()
      • BuilderBase.filter_batch()
      • BuilderBase.infer()
      • BuilderBase.infer_one_step()
      • BuilderBase.load_batch()
      • BuilderBase.process_one_step()
      • BuilderBase.save_downstream()
    • ModelBuilderBase
      • ModelBuilderBase.load_batch()
  • pipeline.utils.modelutils.checkpoint_utils module
    • get_last_artifact()
    • get_last_version_dir()
    • get_last_version_dir_from_config()
    • get_training_metrics()
  • pipeline.utils.modelutils.exploration module
    • ParamExplorer
      • ParamExplorer.add_lhcb_text()
      • ParamExplorer.compute_performance_metrics()
      • ParamExplorer.default_step
      • ParamExplorer.get_output_dir()
      • ParamExplorer.get_performance_from_tracks()
      • ParamExplorer.get_tracks()
      • ParamExplorer.load_preprocessed_dataframes()
      • ParamExplorer.plot()
      • ParamExplorer.run_inference()
  • pipeline.utils.modelutils.export module
    • TRTScatterAddOp
      • TRTScatterAddOp.forward()
      • TRTScatterAddOp.symbolic()
    • change_input_index_types()
    • check_onnx_integrity()
    • convert_model_to_fp16()
  • pipeline.utils.modelutils.metrics module
    • compute_classification_efficiency_purity()
    • compute_efficiency_purity()
  • pipeline.utils.modelutils.mlp module
    • make_mlp()
• pipeline.utils.plotutils package
  • pipeline.utils.plotutils.performance_mpl module
    • get_auto_output_dir()
    • plot_edge_performance()
    • plot_loss()
    • plot_metric_epochs()
    • plot_triplet_performance()
  • pipeline.utils.plotutils.plotconfig module
    • column_labels
    • column_ranges
    • configure_matplotlib()
    • partition_to_color
    • partition_to_label
  • pipeline.utils.plotutils.plotools module
    • add_text()
    • get_figs_axes_on_grid()
    • hide_repetitive_labels_in_grid()
    • pad_on_top()
    • save_fig()
    • set_same_y_lim_for_all_axes()
  • pipeline.utils.plotutils.tracks module
    • plot_tracks()
• pipeline.utils.scriptutils package
  • configure_logger()
  • headline()
  • parse_args()
  • pipeline.utils.scriptutils.convenience_utils module
    • get_training_metrics()
  • pipeline.utils.scriptutils.loghandler module
    • configure_logger()
    • headline()
  • pipeline.utils.scriptutils.parser module
    • add_predefined_arguments()
    • parse_args()
• pipeline.utils.tools package
  • pipeline.utils.tools.tarray module
    • array_to_tensor()
    • count_occurences()
    • get_numpy_or_cupy()
    • get_pandas_or_cudf()
    • get_use_cuda_from_dataframe()
    • series_to_array()
    • series_to_tensor()
    • tensor_to_array()
    • tensor_to_cupy_array()
    • to_dataframe()
  • pipeline.utils.tools.tfiles module
    • delete_directory()
    • is_directory_not_empty()
  • pipeline.utils.tools.tgroupby module
    • get_group_indices()
    • get_group_indices_from_group_lengths()
  • pipeline.utils.tools.tstr module
    • MyFormatter
      • MyFormatter.get_value()
    • partial_format()