Back to the main page.
Bug 584 - Hilbert transform with output 'angle' gives output < -pi and > pi
| Status | CLOSED INVALID |
| Reported | 2011-04-18 20:11:00 +0200 |
| Modified | 2011-04-20 14:01:43 +0200 |
| Product: | FieldTrip |
| Component: | preproc |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P1 normal |
| Assigned to: | Jan-Mathijs Schoffelen |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Jan-Mathijs Schoffelen - 2011-04-18 20:11:50 +0200
Created attachment 43 data to reproduce hilbert angle bug Dag Jan-Mathijs, In de bijlage zit een data set waarmee wij de vreemde output geproduceerd hebben. 1. Wij stoppen de data set opnieuw in ft_preprocessing met cfg.bpfreq=[9 11] en cfg.hilbert=’angle’. 2. Voor het eerste van de vier kanalen is de output OK, maar niet voor de andere drie. Groet, Eric
Jan-Mathijs Schoffelen - 2011-04-18 20:17:26 +0200
Tried to reproduce it using this: cfg=[]; cfg.bpfilter = 'yes'; cfg.bpfreq = [9 11]; cfg.hilbert = 'angle'; data = ft_preprocessing(cfg, data_beam_clean); dat=cat(2,data.trial{:}); minangle = min(dat,[],2); maxangle = max(dat,[],2); minangle = -3.1416 -3.1414 -3.1416 -3.1416 >> maxangle maxangle = 3.1416 3.1416 3.1415 3.1416
Jan-Mathijs Schoffelen - 2011-04-18 22:06:39 +0200
by the way, I tested this with Matlab 2009b on my mac