Closed VeritasJoker closed 1 year ago
The problem was caused because of a workaround that I incorrectly used for the copy-embedding command. After copying the files the correct way (only the embeddings), it worked.
Oh ok. @leoniekerken Can you test encoding without this line df["subject"] = df.subject.astype(int)
and see if it still works?
If it does work then we close this issue. If not we will add this line for the next codebase meeting. Thanks!
Yes, it works without this line. Thanks for your help!
Potentially another bug caused by pickle hierarchy: the new 777_electrode_name.pkl loads the subject column as 'object' and not 'int' so when merging with significant electrode list here causes issues.
Should be fixed by adding this
df["subject"] = df.subject.astype(int)
but not sure so putting it here@leoniekerken is testing it out currently.