Does Pandas `concat` method ignore the `axis` parameter, when concatenating a DataFrame with a Series?

Question

I created a Series object.

for index, entry in a_data_frame.iterrows():
    ...

Then I would like to concatenate this series to a new/another data frame. My goal is to build up the new data frame based on some unique recombination of the rows in the previous one.

a_new_frame = pandas.concat((a_new_frame, a_series))

The series will be appended to the end of the columns disregarding the value of the axis parameter. Why?

geryxyz · Accepted Answer · 2022-02-18 23:31:55Z

1

My experiments allow me to assume that Pandas "thinks" series as columns. When I convert a series to a data frame, it will result in a frame with a single column.

a_series.to_frame()

It makes sense to me that I was unable to use (to concatenate) this series ("column") with a data frame as a "row". The simplest solution is to transpose the new data frame before concatenation.

a_series.to_frame().transpose()

a_new_frame = pandas.concat((a_new_frame, a_series.to_frame().transpose()))

edited Feb 18, 2022 at 23:31

answered Feb 16, 2022 at 21:35

geryxyz

1395 bronze badges

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

Comments

Liam Roberts · Accepted Answer · 2022-02-16 22:46:29Z

0

It's hard to tell what your code is actually doing, but if you want to resample a Dataframe or a Series you can use the .sample() method.

series = pd.Series([1,2,3,4,5])
series.sample(len(series))

Output

2    2
0    1
4    2
1    1
3    2
Name: Path Id, dtype: object

answered Feb 16, 2022 at 22:46

Liam Roberts

1,35110 silver badges7 bronze badges

4 Comments

geryxyz Over a year ago

Thanks for the tip. But no. In a nutshell, I store activities in the data frame, like doing something during my workday and so on. I have to merge subsequent activities of the same type. My time tracking system breaks up these into several instances. But I thought it is irrelevant for the problem mentioned in the question, building up a frame series by series.

Liam Roberts Over a year ago

It would be useful to post the expected input and output you're trying to get. Its hard to tell what the variables are actually representing in the sample code you posted.

geryxyz Over a year ago

Thanks for the remark. I changed the question, but I made it more general because it is independent of the actual problem. The lesson is that pandas "thinks" series as columns by default, not as rows, as I naively thought. I hope I will spare some time for the next one who tries to build a frame similarly.

Liam Roberts Over a year ago

It’s not a default. Series are always used to represent columns in pandas. Any column in a Dataframe is a Series. Glad you managed to figure out a working solution though.

Collectives™ on Stack Overflow

Does Pandas `concat` method ignore the `axis` parameter, when concatenating a DataFrame with a Series?

2 Answers 2

Comments

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related