Selecting columns by name in Pandas - Cheat Sheet & Quick Revision

Recall & Review

beginner

How do you select a single column named 'Age' from a pandas DataFrame?

Use df['Age'] to select the column named 'Age' from the DataFrame df. This returns a Series with the data from that column.

beginner

How can you select multiple columns, for example 'Name' and 'Age', from a pandas DataFrame?

Use df[['Name', 'Age']] to select multiple columns. This returns a new DataFrame with only those columns.

beginner

What happens if you try to select a column name that does not exist in the DataFrame?

pandas will raise a KeyError because the column name is not found in the DataFrame.

intermediate

How do you select columns by name using the .loc accessor?

Use df.loc[:, ['Name', 'Age']] to select columns by name. The colon : means select all rows, and the list specifies the columns.

intermediate

Can you select columns by name using dot notation like df.Age? When is this recommended?

Yes, you can use df.Age to select a column named 'Age'. However, this works only if the column name is a valid Python identifier and does not conflict with DataFrame methods. It's safer to use df['Age'].

Which syntax selects the column 'Salary' from DataFrame df?

A df['Salary', 'Age']

B df.Salary()

C df.loc['Salary']

D df['Salary']

Use df['Salary'] to select a single column by name.

How do you select columns 'Name' and 'Department' together?

A df[['Name', 'Department']]

B df.Name.Department

C df.loc['Name', 'Department']

D df['Name', 'Department']

Use a list inside double brackets: df[['Name', 'Department']].

What does df.loc[:, ['Age', 'Salary']] do?

A Selects all rows and columns 'Age' and 'Salary'

B Selects rows 'Age' and 'Salary'

C Selects columns 'Age' and 'Salary' only for first row

D Raises an error

The colon means all rows, and the list selects the columns.

What error occurs if you select a non-existing column like df['Height']?

A ValueError

B TypeError

C KeyError

D IndexError

pandas raises a KeyError if the column is missing.

Which is the safest way to select a column named 'class'?

A df.class

B df['class']

C df.loc['class']

D df.class()

Use df['class'] because 'class' is a Python keyword and dot notation may fail.

Explain how to select one or more columns by name from a pandas DataFrame.

Think about single vs multiple columns and the loc accessor.

What are the risks or errors when selecting columns by name, and how can you avoid them?

Consider what happens if the column name is missing or unusual.