This project provide python code to do regression on trial-by-trial covariates for MEG/EEG data in the source space, with short-time fourier representation.
The algorithm is based on the following paper Yang, Y., Tarr, M. J., & Kass, R. E. (2014). Estimating learning effects: A short-time fourier transform regression model for MEG source localization. In Springer Lecture Notes on Artificial Intelligence: MLINI 2014: Machine learning and interpretation in neuroimaging (in press by Springer).
Author manuscript can be found at http://www.stat.cmu.edu/~kass/papers/YangTarrKass.pdf
The code depends on the mne-python 0.10. Details can be found at http://martinos.org/mne/stable/