Back to the main page.
Bug 1066 - update wiki and/or comments in ft_componentbrowser_old to plot components using new databrowser
| Status | CLOSED FIXED |
| Reported | 2011-10-21 16:38:00 +0200 |
| Modified | 2012-01-31 12:13:14 +0100 |
| Product: | FieldTrip |
| Component: | documentation |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P1 normal |
| Assigned to: | Johanna |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Johanna - 2011-10-21 16:38:54 +0200
1) http://fieldtrip.fcdonders.nl/example/use_independent_component_analysis_ica_to_remove_ecg_artifacts?s[]=component&s[]=analysis still suggests to use ft_componentbrowser rather than ft_databrowser. 2) 'help' of ft_componentbrowser_old suggests to use ft_databrowser with cfg.viewmode = 'template' (but option does not exist). Instead, if you give a component strucutre to ft_databrowser, then no need to specify viewmode, as cfg.viewmode defaults to 'component' anyway.
Johanna - 2011-10-21 16:42:16 +0200
I've fixed part (2).
Johanna - 2011-10-21 16:56:54 +0200
I've modified wiki text (code section) but not the figure.
Johanna - 2011-10-24 11:25:39 +0200
In attempting to re-do the figure to show ft_databrowser rather than ft_component browser, I ran into 2 issues: 1) I could not find the recommended 'datfile.mat' to load 'data' from. It was not included in ArtifactRemoval.zip. Should it be? 2) I could of course re-create the preprocessed data using the example code, but then does this ICA section want the downsampled 'data' or 'data_orig'? I tried both, but the results from 'data' were rubbish, but using 'data_orig' gave something reasonable. 3) Anytime ICA is run, it gives a different result; i.e. no way to replicate exactly the components and figures already on the wiki. I see some similar components (compared to the wiki) from running it once...should I create new figures with the new result I have obtained? (However, this will also affect the rest of the tutorial as well, as to which components match the ECG)....or just leave the 'old' figure to avoid redoing all this? (Or, it would e easy to remake figure if original 'comp' is still around...).