Black Friday Retail Purchase Data

Columns

• User_ID: A unique identifier for each user.
• Product_ID: A unique identifier for each product.
• Gender: The gender of the user, specified as Male (M) or Female (F).
• Age: The age of the user, provided in age ranges.
• Occupation: A masked numerical value representing the user's occupation.
• City_Category: The category of the city where the user resides (e.g., Type B, Type C).
• Stay_In_Current_City_Years: The number of years the user has resided in their current city.
• Marital_Status: A masked numerical value indicating the user's marital status (e.g., Married or Unmarried).
• Product_Category_1: A masked numerical value representing the primary product category.
• Product_Category_2: A masked numerical value representing a secondary product category, if applicable. This column contains significant missing values.
• Product_Category_3: A masked numerical value representing a tertiary product category, if applicable. This column contains a large proportion of missing values.
• Purchase: The total purchase amount made by the customer; this is the target variable for prediction.

Distribution

• Total Valid Records: 550,000 for most columns.
• User_ID: Ranges from 1,000,001 to 1,006,040.
• Product_ID: Features 3,631 unique products, with 'P00265242' being the most frequently purchased.
• Gender: Approximately 75% of users are Male and 25% are Female.
• Age: The '26-35' age group accounts for 40% of the users. There are 7 unique age groups in total.
• Occupation: There are 20 unique masked occupation values.
• City_Category: City category 'B' is the most common, representing 42% of the data. There are 3 unique city categories.
• Stay_In_Current_City_Years: Users who have stayed 1 year in their current city comprise 35% of the data. There are 5 unique values for years stayed.
• Marital_Status: Masked values for marital status show that roughly 59% are unmarried (value 0) and 41% are married (value 1).
• Product_Category_1: Contains 20 unique masked categories.
• Product_Category_2: Has 18 unique masked categories, but 32% of values are missing.
• Product_Category_3: Has 16 unique masked categories, but a substantial 70% of values are missing.
• Purchase Amount: The average purchase amount is approximately 9,260, with a standard deviation of 5,020. Purchase amounts range from 12 to 24,000.

Usage

• Predicting customer purchase amounts.
• Developing personalised offer strategies for customers.
• Performing Exploratory Data Analysis (EDA) to understand customer and product trends.
• Conducting data cleaning and preprocessing exercises, particularly concerning missing values and categorical data conversion.
• Creating data visualisations to explore relationships such as Age vs. Purchased, Occupation vs. Purchased, and various Product Categories vs. Purchased.
• Implementing Univariate and Bivariate Analysis focusing on the 'Purchase' column.

Coverage

License

Who Can Use It

• Retail companies and analysts aiming to understand customer buying patterns and optimise sales strategies.
• Data scientists and machine learning engineers building predictive models for sales forecasting and personalisation.
• Business intelligence professionals interested in segmenting customers and products based on purchase behaviour.
• Students and researchers for educational purposes in data analysis, machine learning, and retail analytics.
• Developers of data visualisation tools and dashboards.

Dataset Name Suggestions

• Black Friday Retail Purchase Data
• Customer Purchase Prediction Data
• ABC Pvt Ltd Sales Transaction Data
• E-commerce Purchase Behaviour Study

Attributes