Efficiently iterating over rows in a Pandas DataFrame | Maxime Labonne

[utterances-bot]

Efficiently iterating over rows in a Pandas DataFrame | Maxime Labonne

Never use iterrows and itertuples again!

https://mlabonne.github.io/blog/iterating-over-rows-pandas-dataframe/

[CatChenal]

Nice post!
I just want to point out that the drawback you mention under Itertuples only occurs if the column names are not "valid
Python identifiers" as per Pandas docs, which is not the case here.
With namedtuples, the column names (if valid) become accessors at the row level echoing the series level accession with e.g. df.src_bytes:

# Itertuples
total = []
for row in df.itertuples():
    total.append(row.src_bytes + row.dst_bytes)

[mlabonne]

Hi Cat! Thank you for your comment, you're completely right: I forgot to update this version of the article. :) You can even access them with strings thanks to getattr(), which is super nice.

[rashasalim]

Really nice post!
Thanks!

[mlabonne]

Thank you @rashasalim!