Opret array i SELECT

Dette burde gøre tricket:

SELECT a
     , sum(ab_ct)::int AS ct_total
     , count(*)::int   AS ct_distinct_b
     , array_agg(b || ', ' || ab_ct::text) AS b_arr
FROM  (
    SELECT a, b, count(*) AS ab_ct
    FROM   tbl
    GROUP  BY a, b
    ORDER  BY a, ab_ct DESC, b  -- append "b" to break ties in the count
    ) t
GROUP  BY a
ORDER  BY ct_total DESC;

Returnerer:

• ct_total :samlet antal b pr. a .
• ct_distinct_b :antal forskellige b pr. a .
• b_arr :matrix af b plus frekvens af b , sorteret efter frekvens af b .

Sorteret efter det samlede antal b pr. a .

Alternativt kan du bruge en ORDER BY klausul i det samlede kald i PostgreSQL 9.0 eller nyere. Ligesom:

SELECT a
     , sum(ab_ct)::int AS ct_total
     , count(*)::int   AS ct_distinct_b
     , array_agg(b || ', ' || ab_ct::text ORDER BY a, ab_ct DESC, b) AS b_arr
FROM  (
    SELECT a, b, count(*) AS ab_ct
    FROM   tbl
    GROUP  BY a, b
    ) t
GROUP  BY a
ORDER  BY ct_total DESC;

Kan være tydeligere. Men det er typisk langsommere. Og sortering af rækker i en underforespørgsel fungerer for simple forespørgsler som denne. Mere forklaring:

• Hvordan anvender man ORDER BY og LIMIT i kombination med en aggregeret funktion?