Fluctus provides a single interface for handling oscillatory data. This interface is the class Oscillation, which offers methods for normalizing samples, averaging them, interpolating them to a different grid, discarding initial transient timepoints and computing a frequency spectrum. It also offers methods for estimating phase and amplitude, and plotting the data.

This package is WIP and the interface and calling conventions are expected to change as I learn what feels clumsy or confusing.

from fluctus.interfaces import Oscillation
osc = Oscillation(tr=1.0, period=20.0, stimulus_offset=14.0)
# Interpolate to a 100ms grid and percent signal change normalize
osc.interp(target_sampling_out=0.1).psc()
# See all operations that were done to the data
print(osc.transformation_chain)
# Get the transformed data
osc_preproc = osc.transformed_data

Since my use case is dealing with neuroimaging data, it also provides methods for reading data from nifti files given a mask.

osc = Oscillation.from_nifti("mask.nii.gz", "data.nii.gz", period=20.0)

Other utility classes in this package are models for generating HRFs and convolving responses.

fluctus is distributed on PyPI as a universal wheel and is available on Linux/macOS and Windows and supports Python 3.7+.

$ pip install fluctus

The optinal WIP viz module requires ants, which is not listed as a requirement.

fluctus is distributed under the terms of both

• MIT License
• Apache License, Version 2.0

at your option.