elasticai.preprocessor.normalization.normalization

Module Contents

Classes

DataNormalization

API

class elasticai.preprocessor.normalization.normalization.DataNormalization(method: str, do_global_scaling: bool = False, peak_mode: int = 2)

Initialization

Normalizing the input data to enhance classification performance. Parameters: method (str): The normalization method [“minmax”, “norm”, “zscore”, “medianmad”, or “meanmad”] do_global_scaling (bool): Applied global scaling in normalization else sample scaling peak_mode (int): Mode for taking peak value (0: max, 1: min, 2: abs-max) Methods: normalize(): Normalize the input data based on the selected mode and method. Examples: # Create an instance of DataNormalization handler = DataNormalization(“minmax”) data_in = (0.5 - np.random.rand(100, 10)) * 10 normalized_frames = handler.normalize(data_in)

list_normalization_methods(print_output: bool = True) → list

Printing all available methods for normalization

get_peak_amplitude_values() → numpy.ndarray | torch.Tensor

Getting the peak amplitude of rawdata as array

normalize(dataset: numpy.ndarray | torch.Tensor) → numpy.ndarray | torch.Tensor

Apply normalization methods on input data Args: Numpy array with frames for normalizing Returns: Numpy array with normalized frames

create_design(target: str, bitwidth: int, id: str, path2save: pathlib.Path, signed: bool = True) → None

Generate a C design for the configured normalization method.