Performs penalized (and non-penalized) independent component analysis for univariate functional data. Two alternative versions are implemented, both based on the spectral decomposition of the kurtosis operator. Our methods are interfaced with the basis functions provided in the fda package.
❗ For issues, bugs, please use the Github Issues.
The development version of pfica can be installed using:
# install.packages("devtools")
devtools::install_github("m-vidal/pfica")
See the reference manual for more detailed information about the pfica functions.
v0.1.3 beta 25.12.2022
Vidal, M. and Aguilera, A. M. (2022). Novel whitening approaches in functional settings. Stat. 2022;e516, <DOI: 10.1002/sta4.516>.
Vidal, M., Rosso, M. and Aguilera, A. M. (2021). Bi-Smoothed Functional Independent Component Analysis for EEG Artifact Removal. Mathematics 9(11) 1243, <DOI: 10.3390/math9111243>.
ACKNOWLEDGMENTS. S.N Daniel Gost and PAIDI Research Group FQM-307 (UGR).