saftig 0.2.0

Implementations of different static and adaptive filtering techniques for the prediction of a correlated signal component from witness signals.

saftig – Static & Adaptive Filtering Techniques In Gravitational-wave-research

Python implementations of different static and adaptive filtering techniques for the prediction of a correlated signal component from witness signals. The main goal is to provide a unified interface for the different filtering techniques.

Documentation

Features

Static:

• Wiener Filter (WF)

Adaptive

• Updating Wiener Fitler (UWF)
• Least-Mean-Squares Filter (LMS)

Non-Linear:

• Experimental non-linear LMS Filter variant (PolynomialLMS)

Install

From pypi: pip install saftig

From repository: pip install .

From repository (editable): make ie

Minimal example

>>> import saftig as sg
>>>
>>> # generate data
>>> n_channel = 2
>>> witness, target = sg.eval.TestDataGenerator([0.1]*n_channel).generate(int(1e5))
>>>
>>> # instantiate the filter and apply it
>>> filt = sg.filt.LMSFilter(n_filter=128, idx_target=0, n_channel=n_channel)
>>> filt.condition(witness, target)
>>> prediction = filt.apply(witness, target) # check on the data used for conditioning
>>>
>>> # success
>>> sg.eval.RMS(target-prediction) / sg.eval.RMS(prediction)
0.08221177645361015

Terminology

• Witness signal w: One or multiple sensors that are used to make a prediction
• Target signal s: The goal for the prediction