Back to the main page.
Bug 2975 - ft_applymontage does not deal with timelocked data with keeptrials='yes'
| Status |
ASSIGNED |
| Reported |
2015-10-01 14:41:00 +0200 |
| Modified |
2015-10-01 14:44:37 +0200 |
| Product: |
FieldTrip |
| Component: |
forward |
| Version: |
unspecified |
| Hardware: |
All |
| Operating System: |
All |
| Importance: |
P5 minor |
| Assigned to: |
Johanna |
| URL: |
|
| Tags: |
|
| Depends on: |
|
| Blocks: |
|
| See also: |
|
Johanna - 2015-10-01 14:41:39 +0200
Hi all,
If you enter into ft_applymontage data that is the output of ft_timelockanalysis with keeptrials='yes'
e.g.
ans =
avg: [63x3770 double]
var: [63x3770 double]
time: [1x3770 double]
dof: [63x3770 double]
label: {63x1 cell}
trial: [58x63x3770 double]
trialinfo: [58x29 double]
cfg: [1x1 struct]
dimord: 'rpt_chan_time'
then the inputtype is found as 'raw' (due to the presence of a trial field) but then later crashes due to that data.trial is not a cell array, but instead a 3D matrix.
It's an easy fix that I'm happy to do, but which is preferred?
A) Allow this data to be found as 'raw' and then in the switch/case 'raw' section (around line 454) add this line to convert to raw:
data=ft_checkdata(data,'datatype','raw');
or
B) Detect data as timelock and not allow it?
or
C) Detect data as timelock and create new inputtype 'tlock' and switch/case section for this? (with both keeptrials=yes and keeptrials=no options allowed?)
Cheers,
Johanna