Python renaming Pandas DataFrame Columns

Question

import pandas as pd import numpy as np datain = np.loadtxt(datafile) df = pd.DataFrame(data = datain, columns = ["t","p","x","y","z"]) avg = df.groupby(["t"], sort=False)["p"].mean().rename(columns={1:"mean"})

This doesn't work, it tells me TypeError: rename() got an unexpected keyword argument "columns". It also doesn't work if I do this,

avg.rename(columns = {1:"mean"}, inplace=True)

I cannot figure out why, all documentation tells me that my columns call is correct. I just want to rename the blank column created by my "mean" call to have a string index. Anyone know why or how to fix this? All examples I've seen follow this format. Thanks.

Have you tried reading the file in directly with pandas...pd.read_csv(datafile, delimiter = '\t') or similar? — mauve
– mauve, Commented Feb 27, 2019 at 19:26

rpanai · Accepted Answer · 2019-02-27 19:09:15Z

IIUC you could do this

import pandas as pd df = pd.DataFrame({"a":np.arange(10), "b":np.random.choice(["A","B"],10)}) avg = df.groupby("b", sort=False)["a"].mean()\ .reset_index(name="mean")

or

avg = df.groupby("b", sort=False)["a"].mean().reset_index()\ .rename(columns={"a":"mean"})

or

avg = df.groupby("b", sort=False, as_index=False)["a"].mean()\ .reset_index()\ .rename(columns={"a":"mean"})

This worked like a charm, the middle method seemed the cleanest and most straightforward to read to me. Thanks.
It's my personal favorite too. But I wanted to write down few options.

Anna · Accepted Answer · 2020-06-26 02:16:20Z

I ran into this same problem and was also confused about what the issue was. When you call:

df.groupby(...)["p"]....rename(columns={1:"mean"})

the rename() is called on DataFrame["p"] which returns a Series object, not a DataFrame object. The rename() function for a Series object has no column parameter (because there's only 1 "column"). Sometimes, pandas will implicitly convert Series objects to DataFrames so its easy to miss. You could alternatively write

pd.Series.to_frame(df.groupby(...)["p"].mean().reset_index(), name='mean')

Gonzalo Garcia · Accepted Answer · 2019-12-16 15:34:26Z

2

I think this should work:

avg = df.groupby(["t"], sort=False)["p"].mean().rename('mean').reset_index()

edited Dec 16, 2019 at 15:34

Gonzalo Garcia

6,6724 gold badges33 silver badges33 bronze badges

answered Feb 27, 2019 at 19:27

kina_re

411 bronze badge

1 Comment

Will Over a year ago

This gives me TypeError: 'str' object is not callable ... I'm unsure why as I don't fully understand the way rename and reset_index work.

RiveN · Accepted Answer · 2021-11-03 22:34:42Z

I think the problem comes from the fact that when you called:

avg = df.groupby("b", sort=False)["a"].mean().reset_index().rename(columns={"a":"mean"})

This line:

avg = df.groupby("b", sort=False)["a"].mean().reset_index()

returns a pd.Series, not a pd.DataFrame. Normally if you drop the parameters of the column it should work:

avg = df.groupby("b", sort=False)["a"].mean().reset_index().rename("mean")

Collectives™ on Stack Overflow

Python renaming Pandas DataFrame Columns

4 Answers 4

2 Comments

Comments

1 Comment

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

2 Comments

Comments

1 Comment

Comments

Linked

Related