Q1: If I can ask another question here : the dataset i use contains eeg_O1: EEG channel O1-M2, and eeg_O2: EEG channel O2-M1. with their coresponded labels : labels_O1 and labels_O2 (length of the arrays is the same as for other signals, i.e. there is a label per sample.)
I created raw_01 and raw_02 for both of them , what if I want to have one raw object for both how can I stack the signals together with their coresponded labels , does that make sense?
Q2 and if the eeg_O1 for instance is a list of elemnts in different shapes, Is it a good idea to zero padding the elemts so they have the same shape before converting the list to numpy array ?
I’ve moved this question to a new topic thread of its own. In general it is preferred to ask new questions in new threads, not to ask “follow-up” questions that are only marginally related to the original question asked. @Maram Please feel free to edit the title to better reflect what you’re trying to solve.
If I’m understanding your use case correctly, then raw1.add_channels([raw2]) should work. I’m not sure what to say about the “labels” though — are those annotations?
We don’t have enough information about your data to answer that question. Where is the data coming from? Was it previously processed by another program? Why is it in a list with unequal lengths of elements — what is each element, a trial? a subject?
Thanks for your reply.
yes the labels are the annotations. this is the data in this link .
why it is in list : because originally the data are in .mat format , and MNE does not support this format , so I had to use scipy to read the data , then extract the meta data like eeg data, corresponding labels , sampling frequency …etc to create raw object by MNE. You got it ?