Back to the main page.
Bug 65 - cfg.headerformat and dataformat should be supported in all ft main functions that read datathat
| Status | CLOSED WORKSFORME |
| Reported | 2010-04-13 10:03:00 +0200 |
| Modified | 2011-01-26 15:38:31 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P5 normal |
| Assigned to: | Tim Engelkes |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2010-04-13 10:03:21 +0200
I found another problem regarding my data following the above suggestions. Although ft_preprocessing can work well with the 32-bit data with the manual input, ft_definetrial and ft_artifact_eog (and other reading related functions) are not compatible with the 32-bit NeuroScan processing. For example, in line 105 of ft_artifact_zvalue.m, the read_header() function was evoked without the 'headerformat' parameter. There are several other places with the same problem, I listed what I can find below: line 50, trialfun_general.m - read_header(), check if headerformat is provided line 59, trialfun_general.m - read_event(), check if headerformat is provided line 105 & 1152, read_event.m - the reading of neuroscan data is based on a new parameter called eventformat, which is not connected to the headerformat line 149, ft_artifact_zvalue.m - check if headerformat and dataformat are provided line 661, read_data - if the field of dataformat already exists, do not override it I cannot guarantee that all the places were found, but at least my data can be loaded correctly if the above places were fixed.
Tim Engelkes - 2010-05-06 16:08:37 +0200
I have checked the processing of Neuroscan 32 *.cnt data and works now fine for below mentioned functions. Perhaps some additions to keep specified headerformats resolved the problem. The part with eventformat is still not fixed and unclear to me. (In reply to comment #0) > I found another problem regarding my data following the above suggestions. > Although ft_preprocessing can work well with the 32-bit data with the manual > input, ft_definetrial and ft_artifact_eog (and other reading related > functions) are not compatible with the 32-bit NeuroScan processing. > > For example, in line 105 of ft_artifact_zvalue.m, the read_header() function > was evoked without the 'headerformat' parameter. > There are several other places with the same problem, I listed what I can > find below: > > line 50, trialfun_general.m - read_header(), check if headerformat is provided > line 59, trialfun_general.m - read_event(), check if headerformat is provided > line 105 & 1152, read_event.m - the reading of neuroscan data is based on a > new parameter called eventformat, which is not connected to the headerformat > line 149, ft_artifact_zvalue.m - check if headerformat and dataformat are > provided > line 661, read_data - if the field of dataformat already exists, do not > override it > > I cannot guarantee that all the places were found, but at least my data can > be loaded correctly if the above places were fixed. >
Robert Oostenveld - 2011-01-26 15:37:11 +0100
I closed all bugs that were resolved prior and including 2011-01-25. All resolved bugs should have been discussed by now, therefore we don't want to see them again in the next meeting. Instead, in the next meeting we want to see the recent improvements and fixes.