loc vs iloc mental model in Pandas - Practice Questions

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of loc with label slicing

What is the output of this code snippet using loc with label slicing on a DataFrame?

Pandas

import pandas as pd

df = pd.DataFrame({
    'A': [10, 20, 30, 40],
    'B': [1, 2, 3, 4]
}, index=['a', 'b', 'c', 'd'])

result = df.loc['b':'d', 'A']
print(result)

A

b    20
c    30
d    40
Name: A, dtype: int64

B

b    20
c    30
Name: A, dtype: int64

C

1    20
2    30
3    40
Name: A, dtype: int64

D

20    1
30    2
40    3
Name: A, dtype: int64

💡 Hint

Remember that loc includes the end label in slicing.

The loc indexer uses labels and includes the last label in slicing. Here, slicing from 'b' to 'd' includes rows with labels 'b', 'c', and 'd'.

❓ Predict Output

intermediate

Output of iloc with integer slicing

What is the output of this code snippet using iloc with integer slicing on a DataFrame?

Pandas

import pandas as pd

df = pd.DataFrame({
    'A': [10, 20, 30, 40],
    'B': [1, 2, 3, 4]
}, index=['a', 'b', 'c', 'd'])

result = df.iloc[1:3, 0]
print(result)

A

b    20
c    30
Name: A, dtype: int64

B

b    20
c    30
d    40
Name: A, dtype: int64

C

1    20
2    30
Name: A, dtype: int64

D

20    1
30    2
Name: A, dtype: int64

💡 Hint

Remember that iloc uses integer positions and excludes the end index in slicing.

The iloc indexer uses integer positions and excludes the last index in slicing. Here, rows at positions 1 and 2 are selected, which correspond to labels 'b' and 'c'.

❓ data_output

advanced

Difference in output between loc and iloc with mixed index

Given this DataFrame with a non-integer index, what is the output of df.loc[2:4, 'X'] and df.iloc[2:4, 0]?

Pandas

import pandas as pd

df = pd.DataFrame({
    'X': [100, 200, 300, 400, 500],
    'Y': [5, 4, 3, 2, 1]
}, index=[1, 2, 3, 4, 5])

loc_result = df.loc[2:4, 'X']
iloc_result = df.iloc[2:4, 0]
print('loc_result:\n', loc_result)
print('iloc_result:\n', iloc_result)

A

loc_result:
1    100
2    200
3    300
Name: X, dtype: int64
iloc_result:
2    200
3    300
Name: X, dtype: int64

B

loc_result:
2    200
3    300
4    400
Name: X, dtype: int64
iloc_result:
2    200
3    300
Name: X, dtype: int64

C

loc_result:
2    200
3    300
4    400
Name: X, dtype: int64
iloc_result:
3    300
4    400
Name: X, dtype: int64

D

loc_result:
3    300
4    400
5    500
Name: X, dtype: int64
iloc_result:
1    100
2    200
Name: X, dtype: int64

💡 Hint

Recall that loc uses labels and includes the end label, while iloc uses integer positions and excludes the end index.

loc[2:4] selects rows with labels 2, 3, and 4. iloc[2:4] selects rows at integer positions 2 and 3, which correspond to labels 3 and 4.

🔧 Debug

advanced

Identify the error in using loc with integer index

What error will this code raise and why?

Pandas

import pandas as pd

df = pd.DataFrame({
    'A': [1, 2, 3],
    'B': [4, 5, 6]
}, index=[10, 20, 30])

result = df.loc[1]
print(result)

A TypeError: unhashable type: 'int'

B IndexError: single positional indexer is out-of-bounds

C KeyError: 1

D No error, prints the first row

💡 Hint

Check if the label 1 exists in the DataFrame index.

The DataFrame index labels are 10, 20, 30. Using loc[1] looks for label 1, which does not exist, causing a KeyError.

🚀 Application

expert

Selecting rows with loc vs iloc in a real-world scenario

You have a DataFrame df with a date index from '2023-01-01' to '2023-01-10'. You want to select data from '2023-01-03' to '2023-01-06'. Which option correctly selects these rows?

Pandas

import pandas as pd

dates = pd.date_range('2023-01-01', '2023-01-10')
df = pd.DataFrame({'value': range(10)}, index=dates)

# Select rows here

A df.iloc[2:6]

B df.iloc['2023-01-03':'2023-01-06']

C df.loc[2:5]

D df.loc['2023-01-03':'2023-01-06']

💡 Hint

Remember that loc uses labels (dates here), and iloc uses integer positions.

loc selects rows by label, so slicing with date strings works. iloc expects integer positions, so slicing with date strings causes an error.