Setting columns as MultiIndex in Pandas - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of this code when setting MultiIndex columns?

Consider the following pandas DataFrame and code that sets MultiIndex columns. What will be the output of print(df.columns)?

Pandas

import pandas as pd

cols = [('A', 'foo'), ('A', 'bar'), ('B', 'baz')]
df = pd.DataFrame([[1, 2, 3], [4, 5, 6]], columns=pd.MultiIndex.from_tuples(cols))
print(df.columns)

A Index(['A', 'foo', 'A', 'bar', 'B', 'baz'], dtype='object')

B MultiIndex([('A', 'foo'), ('A', 'bar'), ('B', 'baz')], )

C MultiIndex([('foo', 'A'), ('bar', 'A'), ('baz', 'B')], )

D Index(['foo', 'bar', 'baz'], dtype='object')

💡 Hint

Look at how pd.MultiIndex.from_tuples creates a MultiIndex from a list of tuples.

The columns are set as a MultiIndex with levels 'A' and 'B' as the first level and 'foo', 'bar', 'baz' as the second level. So the output shows a MultiIndex with tuples as column labels.

❓ data_output

intermediate

How many columns does the DataFrame have after setting MultiIndex columns?

Given the code below, how many columns does the DataFrame df have?

Pandas

import pandas as pd

arrays = [['one', 'one', 'two', 'two'], ['A', 'B', 'A', 'B']]
cols = pd.MultiIndex.from_arrays(arrays, names=['level_1', 'level_2'])
df = pd.DataFrame([[1, 2, 3, 4], [5, 6, 7, 8]], columns=cols)
print(len(df.columns))

A 2

B 1

C 8

D 4

💡 Hint

Count the number of tuples in the MultiIndex columns.

The MultiIndex columns have 4 tuples: ('one', 'A'), ('one', 'B'), ('two', 'A'), ('two', 'B'). So the DataFrame has 4 columns.

🔧 Debug

advanced

What error does this code raise when setting MultiIndex columns incorrectly?

What error will this code raise?

import pandas as pd
cols = [('A', 'foo'), ('B',)]
df = pd.DataFrame([[1, 2], [3, 4]], columns=pd.MultiIndex.from_tuples(cols))

A ValueError: All tuples must be the same length

B TypeError: Cannot create MultiIndex from tuples

C IndexError: tuple index out of range

D No error, runs successfully

💡 Hint

Check if all tuples have the same length when creating MultiIndex.

MultiIndex requires all tuples to have the same length. Here, one tuple has length 2, the other length 1, so it raises a ValueError.

❓ visualization

advanced

Which option correctly creates a DataFrame with MultiIndex columns and displays it?

Which code snippet will create a DataFrame with MultiIndex columns and display it correctly?

A

import pandas as pd
cols = pd.MultiIndex.from_arrays([['X', 'X', 'Y'], ['a', 'b']])
df = pd.DataFrame([[1,2,3],[4,5,6]], columns=cols)
print(df)

B

import pandas as pd
cols = pd.Index([('X', 'a'), ('X', 'b'), ('Y', 'a')])
df = pd.DataFrame([[1,2,3],[4,5,6]], columns=cols)
print(df)

C

import pandas as pd
cols = pd.MultiIndex.from_tuples([('X', 'a'), ('X', 'b'), ('Y', 'a')])
df = pd.DataFrame([[1,2,3],[4,5,6]], columns=cols)
print(df)

D

import pandas as pd
cols = [('X', 'a'), ('X', 'b'), ('Y', 'a')]
df = pd.DataFrame([[1,2,3],[4,5,6]], columns=cols)
print(df)

💡 Hint

Check how MultiIndex is created and that arrays have matching lengths.

Option C correctly creates a MultiIndex from tuples of equal length and assigns it as columns. Option C uses a list of tuples but not as MultiIndex, so columns are treated as normal tuples, causing error. Option C uses pd.Index with tuples, which does not create MultiIndex. Option C uses from_arrays with arrays of different lengths, causing error.

🚀 Application

expert

After setting MultiIndex columns, what is the value at ('B', 'two') for row 1?

Given the DataFrame below, what is the value at the column ('B', 'two') for the second row (index 1)?

Pandas

import pandas as pd

arrays = [['A', 'A', 'B', 'B'], ['one', 'two', 'one', 'two']]
cols = pd.MultiIndex.from_arrays(arrays)
data = [[10, 20, 30, 40], [50, 60, 70, 80]]
df = pd.DataFrame(data, columns=cols)
value = df.loc[1, ('B', 'two')]
print(value)

A 80

B 70

C 60

D 40

💡 Hint

Look at the column labels and the row index carefully.

The column ('B', 'two') corresponds to the fourth column. For row index 1, the value is 80.