Adding a Custom Biosignal Feature

Below is an example of how to create and register a new feature by inheriting from the base Transform class from MyoVerse.

This custom feature calculates the variance of the input signal.

We Recommend Implementing any Additions in user_config.py

The user_config.py module is specifically designed for end-users to register and configure their own custom components such as models, features, and filters. This keeps your modifications modular, reduces conflicts with core MyoGestic settings, and simplifies upgrades in the future.

Important

By registering your addition in user_config.py, you ensure that your custom configuration stays separate from core MyoGestic functionality and remains compatible with future updates.

Example Overview

1. Define a new feature class inheriting from Transform (TensorTransform).

2. Implement the core logic in the _apply method.

3. Register the feature so it becomes available globally.

import torch
from myoverse.transforms import Transform
from myoverse.transforms.base import get_dim_index

from myogestic.utils.config import CONFIG_REGISTRY

Step 1: Inherit and define logic

class MyVarianceFeature(Transform):
    """
    A feature that computes the variance of the input signal.

    This transform operates on PyTorch tensors with named dimensions.

    Parameters
    ----------
    dim : str, optional
        The dimension along which to compute variance. Default is 'time'.
    keepdim : bool, optional
        Whether to keep the reduced dimension. Default is False.
    """

    def __init__(self, dim: str = "time", keepdim: bool = False, **kwargs):
        super().__init__(dim=dim, **kwargs)
        self.keepdim = keepdim

    def _apply(self, x: torch.Tensor) -> torch.Tensor:
        """
        Compute the variance of the input signal.

        Parameters
        ----------
        x : torch.Tensor
            The input tensor with named dimensions.

        Returns
        -------
        torch.Tensor
            The variance computed along the specified dimension.
        """
        dim_idx = get_dim_index(x, self.dim)
        names = x.names

        result = torch.var(x.rename(None), dim=dim_idx, keepdim=self.keepdim)

        # Restore dimension names if input had named dimensions
        if names[0] is None:
            return result

        if self.keepdim:
            return result.rename(*names)

        new_names = [n for i, n in enumerate(names) if i != dim_idx]
        if new_names:
            return result.rename(*new_names)

        return result

Step 2: Register the custom feature

CONFIG_REGISTRY.register_feature("My Variance Feature", MyVarianceFeature)

Example Usage

if __name__ == "__main__":
    # Create a small signal for demonstration
    sample_data = torch.tensor([1.2, 2.5, 2.7, 2.8, 3.1])
    sample_data = sample_data.rename("time")

    # Instantiate and apply the feature
    feature_instance = MyVarianceFeature(dim="time")
    variance_value = feature_instance(sample_data)

    print(f"Variance of {sample_data.rename(None).numpy()} = {variance_value.item():.4f}")