Получить значение из строк в одной таблице

Question

Я хочу получить статус NEW и тип DATA, всегда самую последнюю сумму для контракта, которая не является NULL.

 
CREATE TABLE mytable(
    Amount VARCHAR(4) NULL 
    ,Status VARCHAR(5) NOT NULL 
    ,Date DATE NOT NULL 
    ,Type VARCHAR(4) NOT NULL 
    ,MetaKey INTEGER NOT NULL 
); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-01','DATA',6); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-02','DATA',7); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-03','DATA',8); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-04','DATA',9); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-05','DATA',10); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-06','DATA',11); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-07','DATA',12); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-08','DATA',13); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-09','DATA',14); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-10','DATA',15); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-11','DATA',16); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-12','DATA',17); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-13','DATA',18); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-14','DATA',19); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-15','DATA',20); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-16','DATA',21); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-17','DATA',22); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-18','DATA',23); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-19','DATA',24); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-20','DATA',25); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-21','DATA',26); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-22','DATA',27); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-23','DATA',28); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-24','DATA',29); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-25','DATA',30); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES ('150','TRANS','2017-01-26','DATA',31); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES (NULL,'NEW','2017-01-27','DATA',32); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES (NULL,'NEW','2017-01-28','DATA',33); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES (NULL,'NEW','2017-01-29','DATA',34); 
INSERT INTO mytable(Amount,Status,Date,Type,MetaKey) VALUES (NULL,'NEW','2017-01-30','DATA',35); 

Ожидаемый выход являются:

 
+--------+--------+------------+------+---------+ 
| Amount | Status | Date | Type | MetaKey | 
+--------+--------+------------+------+---------+ 
| 150 | TRANS | 2017-01-01 | DATA |  6 | 
| 150 | TRANS | 2017-01-02 | DATA |  7 | 
| 150 | TRANS | 2017-01-03 | DATA |  8 | 
| 150 | TRANS | 2017-01-04 | DATA |  9 | 
| 150 | TRANS | 2017-01-05 | DATA |  10 | 
| 150 | TRANS | 2017-01-06 | DATA |  11 | 
| 150 | TRANS | 2017-01-07 | DATA |  12 | 
| 150 | TRANS | 2017-01-08 | DATA |  13 | 
| 150 | TRANS | 2017-01-09 | DATA |  14 | 
| 150 | TRANS | 2017-01-10 | DATA |  15 | 
| 150 | TRANS | 2017-01-11 | DATA |  16 | 
| 150 | TRANS | 2017-01-12 | DATA |  17 | 
| 150 | TRANS | 2017-01-13 | DATA |  18 | 
| 150 | TRANS | 2017-01-14 | DATA |  19 | 
| 150 | TRANS | 2017-01-15 | DATA |  20 | 
| 150 | TRANS | 2017-01-16 | DATA |  21 | 
| 150 | TRANS | 2017-01-17 | DATA |  22 | 
| 150 | TRANS | 2017-01-18 | DATA |  23 | 
| 150 | TRANS | 2017-01-19 | DATA |  24 | 
| 150 | TRANS | 2017-01-20 | DATA |  25 | 
| 150 | TRANS | 2017-01-21 | DATA |  26 | 
| 150 | TRANS | 2017-01-22 | DATA |  27 | 
| 150 | TRANS | 2017-01-23 | DATA |  28 | 
| 150 | TRANS | 2017-01-24 | DATA |  29 | 
| 150 | TRANS | 2017-01-25 | DATA |  30 | 
| 160 | TRANS | 2017-01-26 | DATA |  31 | 
| 160 | NEW | 2017-01-27 | DATA |  32 | 
| 160 | NEW | 2017-01-28 | DATA |  33 | 
| 160 | NEW | 2017-01-29 | DATA |  34 | 
| 160 | NEW | 2017-01-30 | DATA |  35 | 
+--------+--------+------------+------+---------+ 

Я попытался с крестом применить, но это отнимает много времени. Можно ли обойтись без креста?

Answer 1

Declare @LatestAmount VARCHAR(4)= (select Amount FROM mytable WHERE Date = (SELECT MAX(Date) FROM mytable WHERE Amount IS NOT NULL)) 

SELECT Amount,Status,Date,Type,MetaKey FROM mytable WHERE Status= 'TRANS' AND Type='DATA' 
UNION 
SELECT ISNULL(Amount,@LatestAmount),Status,Date,Type,MetaKey FROM mytable WHERE Status= 'NEW' AND Type='DATA' 
ORDER BY Date 

Answer 2

Вы пробовали что-то вроде этого:

SELECT 
    COALESCE(T1.amount,t2.amount) amount 
    ,T1.Status 
    ,T1.DATE 
    ,T1.Type 
    ,T1.MetaKey 
FROM mytable T1 
LEFT JOIN 
    (
    SELECT TOP 1 amount 
    FROM mytable 
    WHERE TYPE = 'DATA' 
    AND amount IS NOT NULL 
    ORDER BY DATE DESC 
    ) T2 
    ON T2.Amount IS NOT NULL 
    AND T1.amount IS NULL 
;