columns and index attributes in Pandas - Step-by-Step Execution

Concept Flow - columns and index attributes

Create DataFrame

↓

Access .columns attribute

↓

Get list of column names

↓

Access .index attribute

↓

Get row labels (index)

↓

Use or modify columns/index

↓

DataFrame updated or queried

Start with a DataFrame, then use .columns to see or change column names, and .index to see or change row labels.

Execution Sample

Pandas

import pandas as pd

df = pd.DataFrame({
  'A': [1, 2],
  'B': [3, 4]
})

cols = df.columns
idx = df.index

Create a DataFrame with columns A and B, then get its columns and index attributes.

Execution Table

Step	Action	Value of df.columns	Value of df.index	Notes
1	Create DataFrame with columns A and B	Index(['A', 'B'], dtype='object')	RangeIndex(start=0, stop=2, step=1)	DataFrame has 2 columns and default index 0,1
2	Access df.columns	Index(['A', 'B'], dtype='object')	RangeIndex(start=0, stop=2, step=1)	Columns attribute shows column names
3	Access df.index	Index(['A', 'B'], dtype='object')	RangeIndex(start=0, stop=2, step=1)	Index attribute shows row labels
4	Modify df.columns to ['X', 'Y']	Index(['X', 'Y'], dtype='object')	RangeIndex(start=0, stop=2, step=1)	Column names changed
5	Modify df.index to [10, 20]	Index(['X', 'Y'], dtype='object')	Int64Index([10, 20], dtype='int64')	Row labels changed
6	Access df.columns and df.index after changes	Index(['X', 'Y'], dtype='object')	Int64Index([10, 20], dtype='int64')	Attributes reflect updates
7	End of trace	Index(['X', 'Y'], dtype='object')	Int64Index([10, 20], dtype='int64')	Execution complete

💡 All steps executed, showing how columns and index attributes are accessed and modified.

Variable Tracker

Variable	Start	After Step 1	After Step 4	After Step 5	Final
df.columns	N/A	Index(['A', 'B'], dtype='object')	Index(['X', 'Y'], dtype='object')	Index(['X', 'Y'], dtype='object')	Index(['X', 'Y'], dtype='object')
df.index	N/A	RangeIndex(start=0, stop=2, step=1)	RangeIndex(start=0, stop=2, step=1)	Int64Index([10, 20], dtype='int64')	Int64Index([10, 20], dtype='int64')

Key Moments - 3 Insights

Why does df.columns show an Index object instead of a simple list?

df.columns returns a pandas Index object which behaves like a list but has extra features; see execution_table rows 1 and 2 where df.columns is shown as Index(['A', 'B'], dtype='object').

What happens if I assign a list of different length to df.columns?

Assigning a list with length different from the number of columns will cause an error; in the trace, step 4 changes columns with a list of matching length, which works fine.

How does changing df.index affect the DataFrame?

Changing df.index changes the row labels; see step 5 where index changes from default RangeIndex to Int64Index with values [10, 20].

Visual Quiz - 3 Questions

Look at the execution_table at step 3, what is the value of df.index?

A Int64Index([10, 20], dtype='int64')

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

C RangeIndex(start=0, stop=2, step=1)

D None

💡 Hint

Check the 'Value of df.index' column at step 3 in execution_table.

At which step does df.columns change to ['X', 'Y']?

A Step 2

B Step 4

C Step 5

D Step 6

💡 Hint

Look at the 'Value of df.columns' column in execution_table to find when the change happens.

If you assign df.index = [0, 1, 2], what would happen?

A It will raise an error because length does not match number of rows

B It will ignore the assignment

C It will work fine and change index to [0, 1, 2]

D It will reset columns instead

💡 Hint

Refer to key_moments about assigning index with different length than rows.

Concept Snapshot

DataFrame.columns and DataFrame.index are attributes to access or set column names and row labels.
.columns returns an Index object listing column names.
.index returns an Index object listing row labels.
You can assign new lists to them but lengths must match existing dimensions.
Changing these attributes updates DataFrame labels accordingly.

Full Transcript

This visual execution trace shows how to use the columns and index attributes of a pandas DataFrame. We start by creating a DataFrame with columns 'A' and 'B' and default row index 0 and 1. Accessing df.columns returns an Index object listing the column names, and df.index returns the row labels as a RangeIndex. We then modify df.columns to ['X', 'Y'], changing the column names, and df.index to [10, 20], changing the row labels. The trace shows the values of these attributes at each step, illustrating how they reflect the DataFrame's structure. Key moments clarify why columns and index are Index objects, the importance of matching lengths when assigning new labels, and the effect of changing the index. The quiz questions test understanding of these attribute values at different steps and consequences of incorrect assignments. The snapshot summarizes the usage rules for these attributes.

Practice

1. What does the columns attribute in a pandas DataFrame represent?

easy

A. The number of rows in the DataFrame

B. The data types of each column

C. The names of the columns in the DataFrame

D. The values of the first row

Solution

1. Step 1: Understand the role of columns in pandas

   The columns attribute stores the labels for each column in the DataFrame, like the headers in a table.

2. Step 2: Compare with other attributes

   Other attributes like index store row labels, not column names.

3. Final Answer:

   The names of the columns in the DataFrame -> Option C

4. Quick Check:

   columns = column names [OK]

Hint: Columns attribute holds column names, like table headers [OK]

Common Mistakes:

• Confusing columns with index
• Thinking columns store data types
• Assuming columns count rows

2. Which of the following is the correct way to get the row labels of a pandas DataFrame df?

easy

A. df.index

B. df.columns

C. df.labels

D. df.rows

Solution

1. Step 1: Recall the attribute for row labels

   The index attribute in pandas stores the row labels or indices.

2. Step 2: Verify other options

   columns is for column names; labels and rows are not valid pandas attributes.

3. Final Answer:

   df.index -> Option A

4. Quick Check:

   index = row labels [OK]

Hint: Use df.index to get row labels, not df.columns [OK]

Common Mistakes:

• Using df.columns to get row labels
• Trying non-existent attributes like df.labels
• Confusing rows with index attribute

3. Given the DataFrame df below:

import pandas as pd
df = pd.DataFrame({
  'A': [1, 2],
  'B': [3, 4]
}, index=['x', 'y'])
print(df.columns)

What will be the output?

medium

A. Index(['A', 'B'], dtype='object')

B. Index(['x', 'y'], dtype='object')

C. ['x', 'y']

D. ['A', 'B']

Solution

1. Step 1: Understand the DataFrame creation

   The DataFrame has columns 'A' and 'B' and row indices 'x' and 'y'.

2. Step 2: Check what df.columns returns

   df.columns returns an Index object listing the column names, which are 'A' and 'B'.

3. Final Answer:

   Index(['A', 'B'], dtype='object') -> Option A

4. Quick Check:

   df.columns = Index of column names [OK]

Hint: df.columns shows column names as Index object [OK]

Common Mistakes:

• Confusing columns with index values
• Expecting a list instead of Index object
• Mixing up row labels with columns

4. What is wrong with this code snippet?

import pandas as pd
df = pd.DataFrame({'X': [10, 20], 'Y': [30, 40]})
df.index = ['a', 'b', 'c']

medium

A. The DataFrame must have numeric index only

B. DataFrame columns cannot be renamed this way

C. The code is correct and runs without error

D. The index list length does not match the number of rows

Solution

1. Step 1: Check DataFrame shape

   The DataFrame has 2 rows but the new index list has 3 labels.

2. Step 2: Understand index assignment rules

   When assigning to df.index, the length must match the number of rows exactly.

3. Final Answer:

   The index list length does not match the number of rows -> Option D

4. Quick Check:

   Index length must equal row count [OK]

Hint: Index length must match rows count exactly [OK]

Common Mistakes:

• Assigning index list with wrong length
• Thinking columns rename affects index
• Assuming any list can be index regardless of length

5. You have a DataFrame df with columns ['name', 'age', 'city'] and default numeric index. How do you rename the columns to ['Name', 'Age', 'City'] and set the index to the 'name' column in one step?

hard

A. df.columns = ['Name', 'Age', 'City']; df.set_index('name', inplace=True)

B. df.rename(columns={'name':'Name', 'age':'Age', 'city':'City'}, inplace=True); df.set_index('Name', inplace=True)

C. df.columns = ['Name', 'Age', 'City']; df.index = df['Name']

D. df.rename(index={'name':'Name'}, inplace=True); df.columns = ['Name', 'Age', 'City']

Solution

1. Step 1: Rename columns properly

   Use df.rename(columns=..., inplace=True) to rename columns safely without losing data.

2. Step 2: Set the index to the renamed 'Name' column

   Use df.set_index('Name', inplace=True) to set the index to the 'Name' column after renaming.

3. Final Answer:

   df.rename(columns={'name':'Name', 'age':'Age', 'city':'City'}, inplace=True); df.set_index('Name', inplace=True) -> Option B

4. Quick Check:

   Rename columns then set index by column name [OK]

Hint: Rename columns with rename(), then set_index() by new name [OK]

Common Mistakes:

• Assigning columns list directly without rename()
• Setting index before renaming columns
• Assigning index to a Series without set_index()