Wide-scan efficiency

Return query_id, query_label, and gb_per_partition (gb_scanned divided by partitions_scanned, as a 2-decimal string). Order by the numeric ratio descending, then query_id.

Divide gb_scanned by partitions_scanned in decimal math, then format the ratio.

Sample input

query_runsgb_scanned divided by partitions_scanned approximates bytes read per partition.

query_id	query_label	gb_scanned	partitions_scanned
1	daily_active_users	12	8
2	orders_export	240	8
3	revenue_by_country	3	2
4	event_funnel	85	6
5	late_event_audit	6	1
6	adhoc_select_star	180	8

Expected output

Expected outputBytes scanned per partition, highest first — the wide SELECT-* scans on top.

query_id	query_label	gb_per_partition
2	orders_export	30.00
6	adhoc_select_star	22.50
4	event_funnel	14.17
5	late_event_audit	6.00
1	daily_active_users	1.50
3	revenue_by_country	1.50

Constraints

Force decimal math with gb_scanned * 1.0 / partitions_scanned, then format with printf('%.2f', ...). Order by the numeric ratio DESC, query_id ASC.

Expected skills

Per-partition scan efficiency and decimal-safe ratio formatting.

SQL

AI evaluation

Run the SQL query to inspect preview rows.