pytspl.filters.base_filter

Base filter.

Classes

BaseFilter

Base filter class for designing linear filters.

Module Contents

class pytspl.filters.base_filter.BaseFilter(simplicial_complex: pytspl.simplicial_complex.SimplicialComplex)

Base filter class for designing linear filters.

sc

history

_reset_history()

Reset the history of the filter design.

set_history(filter: numpy.ndarray, f_estimated: numpy.ndarray, frequency_responses: numpy.ndarray, extracted_component_error: numpy.ndarray, filter_error: numpy.ndarray = np.array([])) → None

Set the history of the filter design.

static calculate_error_NRMSE(f_estimated: numpy.ndarray, f_true) → float

Calculate the error of the estimated signal using NRMSE.

static logistic_function(cut_off_frequency: float = 0.01, steep: int = 100) → numpy.ndarray

Compute the logistic function for the given input.

Args:

cut_off_frequency (float): The cut-off frequency. steep (int): The steepness of the logistic function.

Returns:

np.ndarray: The logistic function output.

static power_iteration(P: numpy.ndarray, iterations: int = 50) → numpy.ndarray

Power iteration algorithm to approximate the largest eigenvalue.

Args:

P (np.ndarray): The input matrix. iterations (int): The number of iterations.

Returns:

np.ndarray: The approximated largest eigenvalue.

get_true_signal(f: numpy.ndarray, component: str) → numpy.ndarray

Get the true signal for the component.

Args:

f (np.ndarray): The signal to be filtered. component (str): The component to be extracted.

Returns:

np.ndarray: The true signal.

get_p_matrix(p_choice: str = 'L1') → scipy.sparse.csr_matrix

Get the matrix P for the filter design. The matrix P can be the Laplacian matrix, the lower Laplacian matrix, or the upper Laplacian matrix.

Args:

p_choice (str, optional): The choice of matrix P. Defaults to “L1”. Choose from [‘L1’, ‘L1L’, ‘L1U’].

Raises:

ValueError: Invalid P_choice.

Returns:

csr_matrix: The matrix P.

get_component_coefficients(component: str) → numpy.ndarray

Calculate the component coefficients of the given component using the order of the eigenvectors.

Args:

component (str): Component of the eigendecomposition to return.

Raises:

ValueError: If the component is not one of ‘harmonic’, ‘curl’, or ‘gradient’.

Returns:

np.ndarray: The component coefficients of the simplicial complex for the given component.