Hello,
For investigational purpose, I have applied different time offsets on different epochs with same number of samples and then averaged those epochs. I was expecting an error telling me that it could not work since epochs have different time definition.
However it worked. I noticed that the average has the same time definition as the first epoch
Thus I am guessing the averaging process automatically performed "Standardize>Uniform Epoch Time", and applied the time definition of the first epoch to all other epochs as it is mentioned in the process window.
Is that correct?
If so, how does "applying a time definition to other epochs" work?
Does it cut the part of the epochs that are out of the first epoch's time definition and add zeros to the parts of the epochs that were not defined?
Or does it simply aligns all epochs' starting point?
Thank you in advance for you help
Best,
François
If the different epochs you average have the same dimensions (same number of time samples and same number of channels), it simply uses the time definition of the first file.
The processes that have a "Time window" option are more picky (eg. statistical tests): since they deal with the selection of specific time points, they would complain if they can't select the same number of samples in all the files.
The process "Standardize>Uniform Epoch Time" can help you solve this case: for files with the same number of time samples but different time values, it reads the Time vector of the first input files and apply it to all the other files.
For files with different numbers of samples, these solutions do not apply: you need to re-epoch your data to select the same number of samples for all the files you want to process together.
Hello François,
thank you for your reply.
That is the case. To add some precisions, I'm trying to see the difference between two averages of epochs.
One is the average of the normal epochs, the other is the average of the same epochs but with a different time offset for each epoch.
If for instance my first epoch has a time definition of [-50, 300] ms and the other epochs have a time definition of [-60, 290] ms, will the average change the time definition of the other epochs to [-50, 300] ms ? Thus cancel out the offsets I applied before ?
If for instance my first epoch has a time definition of [-50, 300] ms and the other epochs have a time definition of [-60, 290] ms, will the average change the time definition of the other epochs to [-50, 300] ms ? Thus cancel out the offsets I applied before ?
It would take the data matrices F as they are and average them, ignoring the Time vector.
The time point -50ms of the first epoch would be averaged with the time point -60ms of the second epochs.
1 Like
Okay, so it cancels out the offsets I've performed before.
Thank you for your help!