Back to the main page.
Bug 3018 - ft_connectivity_granger: write documentation page: how NOT to compute Granger causality with spectral factorization techniques
| Status | CLOSED WONTFIX |
| Reported | 2015-11-29 20:04:00 +0100 |
| Modified | 2019-08-10 12:40:52 +0200 |
| Product: | FieldTrip |
| Component: | documentation |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P5 normal |
| Assigned to: | |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | 3224 |
| See also: |
Jan-Mathijs Schoffelen - 2015-11-29 20:04:24 +0100
of late, there's quite a decent amount of people that want to use the non-parametric spectral factorization approach to compute frequency-domain Granger causality, judging from questions asked on this topic on the mailing list. It seems that uninitiated people don't know how to do it properly. Common mistakes seem to be: - use a very limited amount of frequency bins - try to do the factorization on a wavelet-type of spectral decomposition, leading to 1) NaNs in the TFR, and 2) incorrect factorizations due to different data used to estimate different frequencies. - no doubt some other things can easily go wrong. We could make a FAQ which gives our recommended way of computing it, to be able to point people to in the future.
Jan-Mathijs Schoffelen - 2017-01-17 11:17:14 +0100
It's unclear what the status of this one is, but given the fact that we're severely underpowered in terms of people contributing to fixing issues on bugzilla, Robert and JM have decided to close the low-priority bugs for now. This in order to keep the number of open bugs manageable. Feel free to reopen it, if you are willing to move this one forward towards a more proper resolution.
Robert Oostenveld - 2019-08-10 12:34:42 +0200
This closes a whole series of bugs that have been resolved (either FIXED/WONTFIX/INVALID) for quite some time. If you disagree, please file a new issue on https://github.com/fieldtrip/fieldtrip/issues.