myoverse.datasets.utils.DataSplitter.split#
- DataSplitter.split(data)[source]#
Split data into training, testing, and validation sets.
The split is performed by extracting from the middle of the data: 1. First, test data is extracted from the center 2. Then, validation data is extracted from the center of test data
- Parameters:
data (np.ndarray) – Data to split. First dimension is assumed to be samples.
- Returns:
Named tuple with training, testing, and validation arrays.
- Return type:
SplitResult