Dear SPMers,
I have a scripted M/EEG preprocessing pipeline, at the end of which the M/EEG file is converted to images for statistics. When I last used the pipeline 18 months ago it ran fine, but now I'm running into a warning that I can't figure out. I initially thought that I had made a mistake with an earlier preprocessing step, but I still have the M/EEG files that converted easily to images 18 months ago and now they throw the same warning. I have tried using the GUI to convert a single file rather than batch scripting it, and again I get the same warning. I'm using the same version of Matlab (2013a). Something must therefore have changed in spm_eeg_convert2images or a helper script, but the github history suggests that there have been no changes since Oct 2014. I'm out of ideas, and would really appreciate a hand. I'm happy to send over an example M/EEG file so that others can attempt to replicate (mat+dat = 6MB).
The warning only occurs on conversion of the planar gradiometers, whether or not they are still in pairs or combined. It occurs whether or not there are any badchannels marked. The major problem is that it occurs many thousands of times such that image conversion is not complete even after 12 hours. I can "fix" the problem and get some images out by suppressing the warning (warning off MATLAB:TriScatteredInterp:DupPtsAvValuesWarnId), but I am concerned that this might be masking a significant problem and that the images might not be all they should be.
The repeating error message is:
Warning: Duplicate data points have been detected and removed - corresponding
values have been averaged.
> In griddata>linear at 208
In griddata at 128
In spm_eeg_convert2images at 295
In spm_cfg_eeg_convert2images>run_convert2images at 104
In matlabbatch/private/cfg_run_cm at 29
In cfg_util>local_runcj at 1688
In cfg_util at 959
In cfg_ui>MenuFileRun_Callback at 703
In gui_mainfcn at 96
In cfg_ui at 53
The warning details are:
identifier: 'MATLAB:TriScatteredInterp:DupPtsAvValuesWarnId'
state: 'on'
|