How to use comments to pull specific epochs of data from a given channel / record?

[embello2]

Hello! Thanks so much for putting this package together!

I'm interested in analyzing timeseries data in smaller chuncks, or epochs, defined by the interval between two given comments in the labchart file.

using your sdk, is there a simple way for me to sub-select the interval of data between two comments? I have the comments organized in a table, and could potentially pass two comment number identifiers to specify the interval...

Thanks!

[JimHokanson]

The closest we have now is that you can specify a start and a stop time when retrieving data. With some work you can get the times of the comments and then request the data subset. That being said, I haven't used this package as much as the MATLAB one and there are some convenience functions that I can add in. It is not ready yet but I can certainly do something like:
data = chan.get_data(-1,comment_range=[5,10])

whereas currently you need to extract the times and then do:
data = chan.get_data(record_id,start_time=t1,stop_time=t2)

I just wanted to let you know the current option. I'll update this issue when I have the new interface in place.

[fabriceduprat]

Hi

We are doing a similar thing. We manually add comments to indicate the start and end of seizures on EEG recordings (comments are "begin #" (or "start #") and "end #" with index of the seizure. We then automatically get the seizures datetime and duration using the following code:

    # Get seizure from comments if present
    begin_sec = None
    end_sec = None
    for current_comment in filehandle.records[record_id - 1].comments:
        comment_ch = current_comment.channel_
        comment_text = current_comment.text
        if comment_ch == channel1_index:
            if "begin" in comment_text.lower() or "start" in comment_text.lower():
                begin_sec = current_comment.time - start_sec  # Relative time of comment in seconds (from selected data start time)
            elif "end" in comment_text.lower() and begin_sec:
                end_sec = current_comment.time - start_sec
                seizure_duration = end_sec - begin_sec
                if seizure_duration > min_seizure_duration:
                    if seizure_duration < max_seizure_duration:
                        annotations_starts_sec.append(begin_sec)
                        annotations_durations_sec.append(seizure_duration)
                        annotations_text.append("seizure")
                        seizure_start_datetime = block_start_datetime + datetime.timedelta(seconds=round(begin_sec))
                        seizure_start_datetime_formated = seizure_start_datetime.strftime("%d/%m/%y %H:%M:%S")
                        print(f"Seizure #{len(annotations_starts_sec)} added (start {seizure_start_datetime_formated} for {seizure_duration:.1f} s)")
                    else:
                        print(f"Found seizure aborted, duration {seizure_duration} s > max allowed ({max_seizure_duration})")
                else:
                    print(f"Found seizure aborted, duration {seizure_duration} s < min allowed ({min_seizure_duration})")
                begin_sec = None
                end_sec = None

[embello2]

Thanks for the above!

[JimHokanson]

I updated the README file with an example. The get_data method now supports a comment_range input that takes in two comments IDS and returns the data between those two comments. I pushed changes to GitHub and updated PyPI. Please let me know if that is working for you and then I'll close. Thanks.