Aggregate

APPROX_TOP_COUNT

BigQueryBigQuery

Mengembalikan perkiraan elemen paling sering muncul (top frequent) beserta count-nya. Menggunakan algoritma approximate yang sangat efisien untuk dataset besar.

Tipe hasil: ARRAY<STRUCT<value T, count INT64>>Diperbarui: 7 Jan 2026

Syntax

SQL
APPROX_TOP_COUNT(expression, number)

Parameter

expressionanywajib

Kolom atau ekspresi yang akan dicari top frequent values-nya

numberINT64wajib

Jumlah top values yang ingin dikembalikan

Contoh Penggunaan

Top 5 Produk Terlaris

SQL
1SELECT
2  APPROX_TOP_COUNT(product_name, 5) as top_products
3FROM `project.dataset.orders`;

Mencari 5 produk yang paling sering dipesan.

Hasil
top_products
[{value: "iPhone 15", count: 15230},
{value: "MacBook Pro", count: 8450},
{value: "AirPods", count: 7890},
{value: "iPad", count: 6540},
... 1 baris lainnya

Top Search Keywords

SQL
1SELECT
2  top_keyword.value as keyword,
3  top_keyword.count as search_count
4FROM `project.dataset.search_logs`,
5UNNEST(APPROX_TOP_COUNT(search_query, 10)) as top_keyword;

Mencari 10 keyword pencarian paling populer.

Hasil
keywordsearch_count
laptop gaming125000
iphone 1598000
sepatu nike87500
tas branded76000

Top Errors per Service

SQL
1SELECT
2  service_name,
3  APPROX_TOP_COUNT(error_code, 3) as top_errors
4FROM `project.dataset.error_logs`
5WHERE DATE(timestamp) = CURRENT_DATE()
6GROUP BY service_name;

Mencari 3 error paling sering per service.

Hasil
service_nametop_errors
auth-service[{value: "AUTH_001", count: 5230}, ...]
api-gateway[{value: "TIMEOUT", count: 3450}, ...]
Kembali ke BigQuery Functions