#1
Always order your JOINs from largest tables to smallest tables
#2
Always "GROUP BY" by the attribute/column with the largest number of
unique entities/values
#3
Avoid subqueries in WHERE clause
#4
Use Max instead of Rank
#5
Use ‘regexp_like’ to replace ‘LIKE’ clauses
#6
Use ‘regexp_extract’ to replace ‘Case-when Like’
#7
Convert long list of IN clause into a temporary table
#8 Other Tips
● Use approx_distinct() instead of count(distinct) for very large datasets
● Use approx_percentile(metric, 0.5) for median
● Avoid UNIONs where possible
● Use WITH statements vs. nested subqueries
No comments:
Post a Comment