Setting columns order in pandas dataframe

Question

I have code like this with many columns. The first piece of this code is meant to hard-wire the column order for the final result, as the appended dataframe may have a different order.

pd.DataFrame(columns=['foo', 'bar', 'ddd', 'ccc', 'bbb']).append(pd.DataFrame(data))

With newer versions of pandas I cannot use sort while being backwards compatible to run on e.g. google colab which uses an older version of pandas.

How do I maintain the column order specified above, when appending?

This is not a question about this warning which is elsewhere debated in many places:

FutureWarning: Sorting because non-concatenation axis is not aligned. A future version of pandas will change to not sort by default.

To accept the future behavior, pass 'sort=True'.

It's a question about how to maintain the column order while being backward compatible to older versions of pandas.

What is data? A dictionary?

cs95
– cs95

2018-12-29 11:17:21 +00:00
Commented Dec 29, 2018 at 11:17 — cs95
– cs95, Commented Dec 29, 2018 at 11:17
Yes. Apologies for not mentioning.

matanox
– matanox

2018-12-29 11:18:34 +00:00
Commented Dec 29, 2018 at 11:18 — matanox
– matanox, Commented Dec 29, 2018 at 11:18

cs95 · Accepted Answer · 2018-12-29 11:20:01Z

3

For backward compatibility, I would suggest a reindex step before appending.

df = pd.DataFrame(columns=['foo', 'bar', 'ddd', 'ccc', 'bbb'])

df.append(pd.DataFrame(data).reindex(columns=df.columns))

answered Dec 29, 2018 at 11:20

cs95

406k106 gold badges745 silver badges798 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Setting columns order in pandas dataframe

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related