Is there a way to import an edf file represented as bytes

💬 Support & Discussions
1

I am attempting to parse edf files in parallel in spark and have loaded the files into a binaryfile dataframe. using

https://spark.apache.org/docs/3.0.0-preview/sql-data-sources-binaryFile.html

It seems mne.io.read_raw_edf only supports reading the actual files. i have tried using io.BytesIO and io.RawBaseIO and get the following error

TypeError: expected str, bytes or os.PathLike object, not RawIOBase

so i then tried just casting the bytearray form the dataframe as bytes, and get this error
NotImplementedError: Only EDF files are supported by read_raw_edf, got b’’

does anyone have any suggestions on how to get the edfs loaded into spark using the following method?

2

I’m pretty sure this is not easily possible without adapting the reader function to also accept a file-like (bytes-like) object in addition to a string (the file name). Is someone knows how to do that I’d be very interested!

3

we have done this in the past for fif files see

https://github.com/mne-tools/mne-python/blob/main/mne/io/open.py#L41:5

we could add this for EDF

Alex

4

@rla3rd has opened an issue with a feature request on GitHub: