Back to the main page.
Bug 369 - No horizontal scaling possibility in ft_databrowser on noncontinuous data
| Status | CLOSED FIXED |
| Reported | 2011-01-06 14:28:00 +0100 |
| Modified | 2017-01-25 22:19:18 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P1 enhancement |
| Assigned to: | Roemer van der Meij |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Jörn M. Horschig - 2011-01-06 14:28:28 +0100
When browsing through data with predefined trials (e.g. preprocessed data), there is no possibility to zoom in or out, while it is possible when using continuous data. In the case of noncontinuous data, samples outside the current trial can displayed as zeros.
Robert Oostenveld - 2011-01-14 13:42:48 +0100
(In reply to comment #0) > When browsing through data with predefined trials (e.g. preprocessed data), > there is no possibility to zoom in or out, while it is possible when using > continuous data. > In the case of noncontinuous data, samples outside the current trial can > displayed as zeros. What is better: display samples outside the existing data as zeros (flat line) or as nan (i.e. invisible)
Roemer van der Meij - 2012-03-02 15:40:00 +0100
Stealing a few bugs ;)
Roemer van der Meij - 2012-08-01 12:30:48 +0200
Two things: 1) My proposal is to implement this as follows: when zooming in, the trial is divided into segments, which you can then scroll through by left and right arrows, and NaNs are appended to the last segment of the trial to plot in the same scale. When zooming out further then the original trial boundary, NaNs are appended as well. Let's discuss at the ft-meeting whether this is the desired behavior. 2) In continuous mode now, the last segment is plotted as strechted-out data, it would make more sense to me to do append NaNs to plot at the same scale as the rest of the 'segments'.
Roemer van der Meij - 2012-08-01 18:45:36 +0200
Okay, quite some work :), it is now possible to fully zoom in and zoom out of a trial using the regular horizontal scaling buttons/keys. Jorn, could you give this a quick try? Curious to what you think. For now, I made it such as discussed in the ft-meeting. I.e., when zooming in on a trial, the trial is locked, and you cannot continue to the next trial without zooming out first. It wouldn't be very difficult to unlock this, might be something for the next discussion?
Roemer van der Meij - 2012-08-08 17:17:08 +0200
Reopened: change button name 'trialsegment' to 'segment'. Not an entirely trivial change, as opt.trialname, the field where this is taken from, is used for code-selection/execution. Try to find whether this should be changed/seperated into more options, etc.
Roemer van der Meij - 2012-08-16 18:31:43 +0200
Note to self: opt.trialname is used for determining how to process the range from a selection. Specifically, how to use the offset and begsample columns from opt.trlvis. To do: Look into whether this is necessary, as well for the new trial segment type, and implement another opt.xxx, something like trialtype or the like, to be used next to trialname.
Roemer van der Meij - 2012-08-17 14:47:39 +0200
Renamed trialname to trialviewtype, to better reflect it's purpose (couldn't think of a better name, and is not so important). I also cleaned up the usage with offset. It is now consistent irrespective if trialviewtype is trial, segment, or trialsegment. Offset always correctly reflects what was in the original data, instead of being a special case for segmented data.