How to get a value from a Pandas DataFrame and not the index and object type

Question

Say I have the following DataFrame

Letter    Number
A          1
B          2
C          3
D          4

Which can be obtained through the following code

import pandas as pd

letters = pd.Series(('A', 'B', 'C', 'D'))
numbers = pd.Series((1, 2, 3, 4))
keys = ('Letters', 'Numbers')
df = pd.concat((letters, numbers), axis=1, keys=keys)

Now I want to get the value C from the column Letters.

The command line

df[df.Letters=='C'].Letters

will return

2    C
Name: Letters, dtype: object

How can I get only the value C and not the whole two line output?

On an unrelated note, there's a nicer way to contruct your DataFrame :pd.DataFrame({'Letters': letters, 'Numbers': numbers}) — JoeCondron
– JoeCondron, Commented Jun 11, 2015 at 19:15

user343233 · Accepted Answer · 2024-07-14 04:12:45Z

255

df[df.Letters=='C'].Letters.item()

This returns the first element in the Index/Series returned from that selection. In this case, the value is always the first element.

EDIT:

Or you can run a loc() and access the first element that way. This was shorter and is the way I have implemented it in the past.

edited Jul 14, 2024 at 4:12

user343233

1351 silver badge9 bronze badges

answered Jun 11, 2015 at 18:11

tshallenberger

2,8561 gold badge20 silver badges22 bronze badges

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

4 Comments

Alex Over a year ago

I love this method, however I'm getting the warning: FutureWarning: "item" has been deprecated and will be removed in a future version

Anh-Thi DINH Over a year ago

@AlexG: you can use this instead: df[df.Letters=='C'].Letters.iloc[0]. It produces the first element (which is also the unique) in the result series.

Sonic Soul Over a year ago

using loc[:1] still shows index next to the value :(

user78910 Over a year ago

@AlexG and @Sonic Soul : try using df[df.Letters=='C'].Letters.squeeze() instead. This works the same way. :)

EdChum · Accepted Answer · 2015-06-12 15:28:46Z

99

Use the values attribute to return the values as a np array and then use [0] to get the first value:

In [4]:
df.loc[df.Letters=='C','Letters'].values[0]

Out[4]:
'C'

EDIT

I personally prefer to access the columns using subscript operators:

df.loc[df['Letters'] == 'C', 'Letters'].values[0]

This avoids issues where the column names can have spaces or dashes - which mean that accessing using ..

edited Jun 12, 2015 at 15:28

answered Jun 11, 2015 at 18:21

EdChum

397k204 gold badges836 silver badges583 bronze badges

3 Comments

tshallenberger Over a year ago

It's really inconsequential, but in your selection you access the column 'Letters' using the dot notation; df.loc[df.Letters=='C']. If there are spaces in your column names, you should probably be using converters to strip those out, like you would if importing from a CSV or Excel file.

EdChum Over a year ago

@thomas-ato I'll update my answer but I disagree with modding the columns as an additional step unless that is necessary, in this case I agree it makes no difference

Arya Over a year ago

@EdChum.. In this scenarion : how can we handle error: "IndexError: index 0 is out of bounds for axis 0 with size 0 "

nocibambi · Accepted Answer · 2021-05-07 12:31:26Z

5

You can use loc with the index and column labels.

df.loc[2, 'Letters']
# 'C'

If you prefer the "Numbers" column as reference, you can set it as index.

df.set_index('Numbers').loc[3, 'Letters']

I find this cleaner as it does not need the [0] or .item().

edited May 7, 2021 at 12:31

answered Dec 23, 2020 at 11:26

nocibambi

2,5211 gold badge21 silver badges26 bronze badges

2 Comments

T3metrics Over a year ago

This doesn't address the particular issue. If the index is unknown, your code doesn't help.

nocibambi Over a year ago

The second version (setting one column to index) does apply in that case. :)

Lewis · Accepted Answer · 2019-01-22 04:01:03Z

4

import pandas as pd

dataset = pd.read_csv("data.csv")
values = list(x for x in dataset["column name"])

>>> values[0]
'item_0'

edit:

actually, you can just index the dataset like any old array.

import pandas as pd

dataset = pd.read_csv("data.csv")
first_value = dataset["column name"][0]

>>> print(first_value)
'item_0'

edited Jan 22, 2019 at 4:01

answered Jan 22, 2019 at 3:52

Lewis

3373 silver badges12 bronze badges

Comments

Harry Jones · Accepted Answer · 2022-03-16 11:10:09Z

1

I think a good option is to turn your single line DataFrame into a Series first, then index that:

df[df.Letters=='C'].squeeze()['Letters']

answered Mar 16, 2022 at 11:10

Harry Jones

3623 silver badges10 bronze badges

Collectives™ on Stack Overflow

How to get a value from a Pandas DataFrame and not the index and object type

5 Answers 5

4 Comments

3 Comments

2 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

5 Answers 5

4 Comments

3 Comments

2 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related