Building flows — quiz

What makes the profile pane useful?

1. It is faster
2. It shows value distributions and nulls without writing a query
3. It runs the flow
4. It connects to Desktop

Answer: It shows value distributions and nulls without writing a query — Data-quality problems are visible immediately.

While building, Prep operates on:

1. All the data
2. A sample
3. The first row only
4. Metadata only

Answer: A sample — Which is why rare values may not appear until you run the flow.

The cheapest place to filter rows is:

1. The output step
2. The input step
3. A clean step
4. An aggregate step

Answer: The input step — Excluded rows never enter the flow at all.

To combine a folder of monthly CSVs:

1. One input each
2. A wildcard union on the input
3. A join step
4. A pivot step

Answer: A wildcard union on the input — Prep adds a File Paths field recording the source.

The join step shows fewer rows than expected. Best first action?

1. Change join type
2. Click the unmatched section to see which values failed
3. Add a clean step
4. Re-run the flow

Answer: Click the unmatched section to see which values failed — The unmatched values are usually a formatting difference.

The row count after a join exceeds the left input. This means:

1. Normal
2. Duplicate keys on the right caused a fan-out
3. Rows were lost
4. A union is needed

Answer: Duplicate keys on the right caused a fan-out — Aggregate the right branch by the key first.