總結記錄

Question

我有一個像

╔═════╦══════╦══════════╦══════════╗ 
║ PID ║ NAME ║ DATE1 ║ DATE2 ║ 
╠═════╬══════╬══════════╬══════════╣ 
║ 123 ║ john ║ 20110401 ║ 19700101 ║ 
║ 123 ║ john ║ 20110401 ║ 19700101 ║ 
║ 123 ║ john ║ 20110401 ║ 19700101 ║ 
║ 323 ║ mike ║ 20110401 ║ 19900101 ║ 
║ 323 ║ mike ║ 20110401 ║ 19900101 ║ 
╚═════╩══════╩══════════╩══════════╝ 

query1記錄一個表是count上pid

query2是，如果所有的四個字段相同的記錄被視爲一個記錄其他個人記錄和記錄按pid的順序計算。

上述結果是：

╔═════╦══════════════╦═════════════╗ 
║ PID ║ TOTALRECORDS ║ TOTALUNIQUE ║ 
╠═════╬══════════════╬═════════════╣ 
║ 123 ║   3 ║   1 ║ 
║ 323 ║   2 ║   1 ║ 
╚═════╩══════════════╩═════════════╝ 

我做select pid,count(pid),(select count(pid) group by pid,name,date1,date2) from <table> group by pid;

但我得到一個錯誤......好心糾正我的代碼。

Answer 1

SELECT pid, 
     COUNT(*) totalRecords, 
     COUNT(DISTINCT name, date1, date2) totalUnique 
FROM TableName 
GROUP BY pid 

• SQLFiddle Demo
• SQLFiddle Demo (with not unique record)

結果

╔═════╦══════════════╦═════════════╗ 
║ PID ║ TOTALRECORDS ║ TOTALUNIQUE ║ 
╠═════╬══════════════╬═════════════╣ 
║ 123 ║   3 ║   1 ║ 
║ 323 ║   2 ║   1 ║ 
╚═════╩══════════════╩═════════════╝