Weighted Phase Lag Index (Vinck, Oostenveld, van Wingerden, Battaglia & Pennartz 2011) between every channel pair. More...
#include "../connectivity_global.h"#include "abstractmetric.h"#include "../connectivitysettings.h"#include <QSharedPointer>#include <QMutex>#include <Eigen/Core>

Go to the source code of this file.
Classes | |
| class | CONNECTIVITYLIB::WeightedPhaseLagIndex |
| Weighted Phase Lag Index estimator (Vinck et al. 2011); volume-conduction-robust with lower bias than PLI. More... | |
Namespaces | |
| namespace | CONNECTIVITYLIB |
| Functional connectivity metrics (coherence, PLV, cross-correlation, etc.). | |
Weighted Phase Lag Index (Vinck, Oostenveld, van Wingerden, Battaglia & Pennartz 2011) between every channel pair.
SPDX-License-Identifier: BSD-3-Clause Copyright (c) 2018-2026 MNE-CPP Authors
The weighted PLI replaces the unweighted sign average used by PhaseLagIndex with an imaginary-part-magnitude-weighted version,
wPLI_{xy}(f) = | E[ Im(S_{xy}(f)) ] | / E[ |Im(S_{xy}(f))| ]
Vinck et al. (NeuroImage 2011) showed that this estimator dominates the plain PLI on three counts: it is less sensitive to phase fluctuations around the zero-lag line (because near-zero Im(S_{xy}) values contribute proportionally little to both numerator and denominator), it has lower sample-size bias, and it has higher statistical power to detect genuine non-zero lag interactions in noisy MEG/EEG. Like PLI and imaginary coherence, the construction guarantees rejection of zero-lag volume-conduction / common-reference mixing.
The output is bounded in [0, 1] and is still positively biased for small numbers of trials. The squared, debiased variant DebiasedSquaredWeightedPhaseLagIndex - also from Vinck et al. (2011) - removes that small-sample bias analytically and should be preferred when the trial count is low.
Definition in file weightedphaselagindex.h.