1. Data Science vs Data Analytics
1.3 Grain of One Row
Grain (तपशीलाची पातळी) means what one row represents. Wrong grain breaks every total.
Our primary table orders has one row per order line (one product on one order):
| order_id | order_date | city | product | qty | amount | status |
|---|---|---|---|---|---|---|
| BLK-260314-001 | 14-03-2026 | Pune | Gokul Cow Milk 500 ml | 2 | 64 | Delivered |
| BLK-260314-001 | 14-03-2026 | Pune | Ladi Pav (6 pcs) | 1 | 30 | Delivered |
Two rows, same order_id — still line grain. Counting rows ≠ counting orders. Use nunique() on order_id for order count.
Practice task
If Raja loads a sheet where each row is already “daily GMV per city”, can he compute average order value from that sheet alone? Explain in one sentence.