Adding And Removing Columns

Pandas: Adding & Removing Columns

In data analysis, your dataset is rarely static. You often need to add new data (like assigning grades) or clean up unwanted data. In this tutorial, we will learn the different ways to create columns and the proper way to delete them.

---

🔹 1. Adding New Columns

There are two main ways to add a new column to a DataFrame: assigning a list of values or assigning a single value to everyone.

A. Assigning a List (Different Values for Each Row)

If you want every row to have specific data, you assign a list.

🧪 Code Example

import pandas as pd

# Setup sample data
data = {
    "Name": ["Atul", "Aman", "Harsh", "Rishi", "Devansh"],
    "Class": [11, 39, 47, 56, 76],
    "Status": ["pass", "fail", "pass", "pass", "fail"]
}
table = pd.DataFrame(data)

# Add a "Grade" column with specific values
table["Grade"] = ["A", "AB", "C", "D", "E"]
print(table)

📤 Output

      Name  Class Status Grade
0     Atul     11   pass     A
1     Aman     39   fail    AB
2    Harsh     47   pass     C
...

⚠ Important Warning: The list length must match the DataFrame length exactly. If you have 5 rows, you must provide 5 values.

B. Assigning a Single Value (Broadcasting)

If you want to assign the same value to every single row, you just assign that one value. Pandas is smart enough to repeat it for you. This is called "Broadcasting".

🧪 Code Example

# Assign "Indian" to the "Country" column for EVERYONE
table["Country"] = "Indian"
print(table)

📤 Output

      Name  Class Status Grade Country
0     Atul     11   pass     A  Indian
1     Aman     39   fail    AB  Indian
2    Harsh     47   pass     C  Indian
...

---

🔹 2. Removing Columns (The .drop Method)

To delete columns, we use the .drop() method.

A. Temporary Removal (Default Behavior)

By default, .drop() creates a new copy of the data with the column removed. The original table stays untouched.

Why axis=1?

• axis=0 (Default) = Searches for the label in the Rows (Index).
• axis=1 = Searches for the label in the Columns (Headers).

If you forget axis=1, Pandas will look for "Grade" in the row numbers, fail to find it, and give an error.

🧪 Code Example

# Drop "Grade" temporarily
removed_temp = table.drop("Grade", axis=1)

print("----------Temporary Removal-----------------")
print(removed_temp)

print("----------Main Data (Unchanged)-------------")
print(table)

📤 Output

----------Temporary Removal-----------------
      Name  Class Status Country
0     Atul     11   pass  Indian
...

----------Main Data (Unchanged)-------------
      Name  Class Status Grade Country
0     Atul     11   pass     A  Indian
...

---

B. Permanent Removal (Inplace)

If you want to remove the column from the memory permanently, use inplace=True. This modifies the original variable.

🧪 Code Example

# Remove "Grade" forever from the original table
table.drop("Grade", axis=1, inplace=True)

print("----------Main Data (Modified)--------------")
print(table)

📤 Output

----------Main Data (Modified)--------------
      Name  Class Status Country
0     Atul     11   pass  Indian
...

---

🔹 3. Removing Multiple Columns

To remove more than one column, pass a list of column names.

🧪 Code Example

# Remove "Class" and "Status" temporarily
removed_temp = table.drop(["Class", "Status"], axis=1)

print(removed_temp)

📤 Output

      Name Country
0     Atul  Indian
1     Aman  Indian
...

---

📌 Summary

Goal	Syntax Code	Note
Add List	df["Col"] = [1, 2, 3]	Length must match rows.
Add Single Value	df["Col"] = "A"	Fills entire column with "A".
Temp Drop	df.drop("Col", axis=1)	Original data safe.
Perm Drop	... inplace=True	Original data modified.