Create and Transform Features for Insights

Business Scenario

Welcome!

Today is your seventh day as a Junior Data Analyst at a retail analytics company.

Retail organizations often need additional insights that are not directly available in raw datasets. Data analysts create new features (columns) and transform existing data to uncover meaningful business patterns, improve reporting, and support decision-making.

Pre-Lab Preparation

Click here to download previous lab file: DM LAB 6

git pull origin branchName

Git Pull

Click to download Dataset : Retail_Dataset_Cleaned

Topic: Data Manipulation with Pandas

1) Data Preparation Essentials

Task 1: Create New Columns

Retail organizations collect large amounts of sales, inventory, customer, and product data every day. However, raw data alone does not always provide the insights needed for business decision-making.

Data analysts often create new columns (features) from existing data to derive meaningful business metrics. These newly created features help understand profitability, inventory levels, sales performance, and customer behavior more effectively.

By creating new features, analysts can transform raw data into actionable business insights.

What is Feature Creation?

Feature Creation is the process of generating new columns using existing columns in a dataset.

Open Google Colab

1

2

Import Required Libraries

import pandas as pd
import numpy as np

3

Upload the Retail Dataset

4

Load Dataset Using Pandas

df = pd.read_csv("/content/Retail_Dataset_Modified.csv")

print("Dataset Loaded Successfully")

Display First Five Records

5

df.head()

6

Check Dataset Information

df.info()

7

Create Profit Amount Column

Profit Amount indicates the actual profit earned from sales.

Profit Amount = Revenue × Profit Margin / 100

Formula

df["Profit_Amount"] = (
    df["Revenue"] *
    df["Profit_Margin"] / 100
)

df[["Revenue",
    "Profit_Margin",
    "Profit_Amount"]].sample(5)

Create Stock Difference Column

This feature shows the difference between available stock and units sold.

8

df["Stock_Difference"] = (
    df["Quantity_Available"] -
    df["Units_Sold"]
)

df[["Quantity_Available",
    "Units_Sold",
    "Stock_Difference"]].sample(5)

9

8

Create Inventory Status Column

Categorize inventory levels based on available quantity.

Condition:

  • Low Stock → Quantity < 100
  • Medium Stock → Quantity 100–200
  • High Stock → Quantity > 200
df["Inventory_Status"] = np.where(
    df["Quantity_Available"] < 100,
    "Low Stock",
    np.where(
        df["Quantity_Available"] <= 200,
        "Medium Stock",
        "High Stock"
    )
)

df[["Quantity_Available",
    "Inventory_Status"]].sample(5)

Create Revenue Per Unit Column

Revenue generated for each unit sold.

10

df["Revenue_Per_Unit"] = (
    df["Revenue"] /
    df["Units_Sold"]
)

df[["Revenue",
    "Units_Sold",
    "Revenue_Per_Unit"]].sample(5)

Verify Newly Created Columns

11

print(df[[
    "Profit_Amount",
    "Stock_Difference",
    "Inventory_Status",
    "Revenue_Per_Unit"
]].head())

Task 2: Transform Data

Retail datasets often contain raw numerical and textual data that may not be easy to interpret directly.

For example:

  • Revenue values such as 8,452 or 12,340 may be difficult to analyze individually.
  • Customer Satisfaction scores such as 4.5 or 3.2 may not clearly indicate customer sentiment.
  • Discount percentages may need to be grouped into categories for reporting.

 

To make data easier to understand and analyze, analysts perform data transformation.

What is Data Transformation?

Data Transformation is the process of converting data from one format or structure into another format that is more useful for analysis.

Transform Revenue into Revenue Category

1

df["Revenue_Category"] = np.where(
    df["Revenue"] < 5000,
    "Low Revenue",
    np.where(
        df["Revenue"] < 10000,
        "Medium Revenue",
        "High Revenue"
    )
)

df[["Revenue", "Revenue_Category"]].head()

np.where() when we want to create a new column based on a condition.

2

Transform Customer Satisfaction Ratings

df["Customer_Feedback"] = np.where(
    df["Customer_Satisfaction"] >= 4,
    "Satisfied",
    "Needs Improvement"
)

df[[
    "Customer_Satisfaction",
    "Customer_Feedback"
]].head()

3

Create Discount Category

df["Discount_Category"] = np.where(
    df["Max_Discount_Offered"] < 10,
    "Low Discount",
    np.where(
        df["Max_Discount_Offered"] < 20,
        "Medium Discount",
        "High Discount"
    )
)

df[["Max_Discount_Offered", "Discount_Category"]].sample(5)

4

Create Delivery Speed Category

df["Delivery_Speed"] = np.where(
    df["Delivery_Time"] <= 5,
    "Fast Delivery",
    "Standard Delivery"
)

df[[
    "Delivery_Time",
    "Delivery_Speed"
]].head()

Display Transformed Columns

5

print(df[[
    "Revenue_Category",
    "Customer_Feedback",
    "Discount_Category",
    "Delivery_Speed"
]].head())

6

Check Dataset Information

df.info()

 

Great job!

You have successfully completed your lab on Create and Transform Features for Insights. In this lab, you have: Created new columns using existing data, Transformed numerical data into meaningful categories, Created business-friendly labels for reporting, Prepared the dataset for advanced analysis and visualization

You are now ready to move to the next stage of Junior Data Analyst.

Checkpoint

   Git Push

git push origin branchName

Next-Lab Preparation

Topic: Decoding Your Data

1) Unraveling the World of Data and Its Insights

2) The Basics of Statistics: Descriptive Insights
3) Measuring Data Spread: Dispersion Insights

DM7 LAB: Create and Transform Features for Insights

By Content ITV

DM7 LAB: Create and Transform Features for Insights

  • 21

More from Content ITV