Back to the main page.

Bug 1621 - ft_databrowser: viewmode vertical, default vertical scaling does not make sense (too zoomed out)

Reported 2012-08-01 14:24:00 +0200
Modified 2013-01-17 10:40:48 +0100
Product: FieldTrip
Component: plotting
Version: unspecified
Hardware: PC
Operating System: Windows
Importance: P3 trivial
Assigned to: Roemer van der Meij
Depends on:
See also:

Roemer van der Meij - 2012-08-01 14:24:37 +0200

The default scaling, maxabs, does not make sense when viewmode is vertical. Every channel is a almost a flat line and you always have to zoom in before using. I propose to automatically decrease the default scaling when viewmode is vertical, by a factor of 10 (works nicely for me). JM and Jorn, as it's a default, could you give a short no or yes on whether this also works nicely for you? If you input cfg.develscalefix = 'yes' it uses the default I semi-randomly picked. On line 362 you can find the factor I used to adjust if you want to try it out.

Roemer van der Meij - 2012-08-01 14:25:20 +0200

Assigning and CC'ing

Jörn M. Horschig - 2012-08-01 16:02:15 +0200

Not quite sure (didn't test though, yet). Some questions to think about though: - Is it useless when using continuous data or segmented data? Is the scaling for segmented data done on the maxabs of all trials or only the current trial? - Is it for EEG or MEG data (#channels)? I'm gonna check with my (i)EEG and MEG data and let you know. Is this high priority?

Roemer van der Meij - 2012-08-01 16:39:29 +0200

It is of extremely low priority ;). It just annoyed me that I was pressing shift+up a number of times every time I opened a databrowser, and realized I wasn't the only one probably. I updated the priority accordingly ;)

Roemer van der Meij - 2012-08-02 15:09:50 +0200

Minor comment: The scaling is useless whenever the viewmode is vertical I think. It is also not caused by doing a maxabs (which I am talking about) on only part of the data (which is true), the silly issue is caused because there is currently no overlap allowed for channel time-courses, such that each one has it's own white-space to fully use. When viewing time-courses above each other though, it makes sense for them to overlap moderately, otherwise, you mainly see white space, instead of time-courses :) EEG/MEG shouldn't matter, the above would be the same for all. Hell, it would even hold for the social sciences time-series ;). Whether continuous is yes or no doesn't matter as well.

Roemer van der Meij - 2012-08-07 15:43:29 +0200

I went ahead and enabled it by default ;)

Roemer van der Meij - 2013-01-17 10:40:48 +0100

bug closing time