Snowflake Project 11: Data Sampling

Overview

This project focuses on data sampling in Snowflake, demonstrating two methods:

1. Random Sampling
2. System Sampling

The project shows how to create databases, tables, stages, load data from S3, perform sampling, and clean up resources.

---

Tasks

TASK 1: Random Sampling of Data

• Create database: sales_db_random_sampling
• Create table: sales_data
• Load data from S3 stage
• Sample 5% of the data randomly
• Query results to verify sampling
• Cleanup resources

TASK 2: SYSTEM Sampling Method

• Create database: sales_db_system_sampling
• Create table: sales_data
• Load data from S3 stage
• Perform SYSTEM sampling (~5% of data)
• Query results to verify sampling
• Cleanup resources

---

Real-World Relevance

• Random sampling helps in analyzing large datasets efficiently without loading the full dataset.
• SYSTEM sampling is useful for approximate query results with minimal performance overhead.
• Both methods are critical for data exploration, testing, and building sample datasets for analytics pipelines.

---

Tools & Features Covered

• Snowflake Databases and Tables
• External Stages pointing to S3
• COPY command for data loading
• RANDOM and SYSTEM sampling in Snowflake
• Data cleanup best practices