selfeeg.utils

This module simply gathers functions and classes for various purposes. For example, you can find a torch implementation of the Scipy’s pchip interpolation function (used for resampling) or pytorch EEG scaler with a soft clipping option. Both the cited functions are compatible with GPU tensors.

Classes

`RangeScaler`	class adaptation of the `scale_range_with_soft_clip` function.
`ZscoreScaler`	zscore operator callable objects.

Functions

`check_models`	checks that two nn.Modules are equal.
`count_parameters`	counts the number of trainable parameters of a Pytorch's nn.Module.
`create_dataset`	creates a simulated EEG dataset for normal abnormal binary classification.
`get_subarray_closest_sum`	find the subarray whose values sum is closer to a target.
`scale_range_soft_clip`	soft version of the range scaler.
`torch_pchip`	performs the pchip interpolation on the last dimension of the input tensor.
`torch_zscore`	zscore operator for torch tensors.