Practice — Medium · Counting What Matters
Skills focus. Changing a field's summarization (sum → count distinct) ·
bar and line charts · reading a chart for its trough, not just its peak.
Ungraded. In-class or self-paced. Less scaffolding than Mild — the steps say what, you work out how.
Scenario
The operations manager does not care about dollars this morning — she staffs the warehouse and
wants to know about order volume: how many orders came in, from where, and when
the quiet months fall. Same file, different lens: l01_retail_2025.csv.
The one new move
This practice needs a count of orders, and multiple rows can share one
OrderID (one row per product in the order). Summing is wrong and plain counting
double-counts. In the visual's field well, open the dropdown on OrderID and choose
Count (Distinct) — count each order once, however many lines it has.
Tasks
- Load
l01_retail_2025.csvinto a new file. - Build a card showing the number of distinct orders in 2025.
- Build a bar chart of distinct orders by
Region, largest on top. Which region takes the most orders? Which takes the fewest? - Build a line chart of
Revenueby month. Ignore the peak this time — find the trough. Which month is the year's quietest, and roughly how much revenue did it bring in? - Sanity-check yourself: your order-count card and your revenue card from the lesson describe the same year. Divide one by the other in your head — is an average order of roughly $780–$790 believable for a home-goods retailer? (This habit — dividing two big numbers to see if their ratio is sane — will save you repeatedly in this course.)
Hint — my order count looks far too big
A count of 50,539 is the row count — you counted order lines. Open the field's dropdown in the field well and switch from Count to Count (Distinct).
Reflection
- When would a business care about order count rather than revenue? Name two decisions each number drives.
- February is the trough for revenue. Would you bet it is also the trough for order count? How would you check in under a minute?