How to do this in pandas:
I have a function extract_text_features on a single text column, returning multiple output columns. Specifically, the function returns 6 values.
The function works, however there doesn't seem to be any proper return type (pandas DataFrame/ numpy array/ Python list) such that the output can get correctly assigned df.ix[: ,10:16] = df.textcol.map(extract_text_features)
So I think I need to drop back to iterating with df.iterrows(), as per this?
UPDATE:
Iterating with df.iterrows() is at least 20x slower, so I surrendered and split out the function into six distinct .map(lambda ...) calls.
UPDATE 2: this question was asked back around v0.11.0, before the useability df.apply was improved or df.assign() was added in v0.16. Hence much of the question and answers are not too relevant.