I have managed to insert one event into my EEG FIF file, which I want to average (yes, Nave = 1). I did this by importing a text file list of events using mne-browse-raw. I can see this event when I list them within mne-browse-raw. When I try to average inside the GUI, I get a brief display of a window that goes by too fast to read, then the message "nothing accepted"
Any suggestions on debugging this? Some of my files are included below. Thanks,
-Jeff
Here is the event file I created:
# sample time from to text
0 0 0 0 psuedo_event
10 0.009 0 1 trigger1
Here is the ave file I created:
average {
outfile A1_ec1_bcm_epc1-ave.fif
eventfile A1_ec1_bcm_epc1.eve
logfile A1_ave_log.txt
category {
name ave1epoch
event 1
tmin 0.0
tmax 0.9
}
}
Here is the console output of trying to average:
Event file is in the new format
Read 1 events from A1_ec1_bcm_epc1.eve
test : nave = 0
Average :
stim ignore : off
grad reject : off
mag reject : off
EEG reject : off
EOG reject : off
ECG reject : off
grad flat : off
mag flat : off
EEG flat : off
EOG flat : off
ECG flat : off
Do not fix trigger skew
output file : A1_ec1_bcm_epc1-ave.fif
log file : A1_ave_log.txt
event file : A1_ec1_bcm_epc1.eve
test :
t = 0.0 ... 900.0 ms
event = 1
ignore = 0
Event file is in the new format
Read 1 events from A1_ec1_bcm_epc1.eve
test : nave = 0
Which version of MNE are you running? I think that one of the older
versions of MNE didn't recognize the last event in a text file, which
you could overcome by adding a dummy event line like this:
The choice of 111 as the trigger number here is arbitrary, it's just
putting something there so the second-to-last line (important one)
actually gets picked up.
I am using v2.7.3 on a Macintosh.
I verified that it does read the events in properly - if I use your
suggestion I get two events.
I left the single event in, and saved events in a -eve.fif file. Below is
the result of the last mne_browse_raw session. I tried to average using
Process:Average, and all indications seem to confirm the event is in
there, but in the end is says "nave = 0".
I wonder if there are some default rejection settings that are still in
effect and for some reason it is rejecting the epoch based on some
artifact criteria?
I think you might have mentioned this in a previous email, but does it
actually produce the "A1_ave_log.txt" log file? That file should
record any trials that were rejected, but I'd be surprised if
rejections were the issue.
You could try running something like this (assuming the syntax is adjusted):
With the events (with and then without the dummy line I mentioned
before) in a text file named A1_ec1_bcm_epc1.eve, in the directory
where you're calling mne_process_raw. I believe this will cause MNE to
ignore the events in the actual FIF while doing the averaging and
instead use the text events read from the text file.
Also, it just occurred to me that I've always used a "bmin" and "bmax"
to set the baseline when doing averaging---it's possible MNE defaults
to -200ms to 0ms for the baseline if you don't tell it otherwise, in
which case I think your first epoch wouldn't be used because there is
no data for the baseline time span.
I have to go home now and pack for a trip, but will try your suggestion
soon. Now I recall I tried something like it last week, but was not
successful then.