有条件地将where子句添加到MySQL查询

Question

目的

要获取ratings.id s，其中ratings.ticker在日期范围内增加5％：
在哪里prices.price_date > ratings.date_issued
并且，如果（对该股票行情/股票经纪人组合发布了另一个评级），那么prices.price_date < new_rating.date_issued否则不设置小于日期的where子句

当前设置

表：等级

| id   | ticker | broker | issue_date | close_price |
|------|--------|--------|------------|-------------|
| 22   | ABF    | HSBC   | 2017-01-23 | 2523        |
| 775  | ABF    | HSBC   | 2017-02-15 | 2527        |
| 1111 | ABF    | HSBC   | 2017-02-28 | 2619        |

表：价格

| id    | ticker | price_date | price   |
|-------|--------|------------|---------|
| 42    | ABF    | 2017-01-24 | 2533.52 |
| 80    | ABF    | 2017-01-25 | 2531.00 |
| 145   | ABF    | 2017-01-26 | 2500.00 |
| 263   | ABF    | 2017-01-27 | 2420.00 |
| 416   | ABF    | 2017-01-28 | 2440.00 |
| 585   | ABF    | 2017-01-29 | 2440.00 |
| 754   | ABF    | 2017-01-30 | 2440.00 |
| 923   | ABF    | 2017-01-31 | 2449.00 |
| 1112  | ABF    | 2017-02-01 | 2440.00 |
| 1315  | ABF    | 2017-02-02 | 2405.00 |
| 1535  | ABF    | 2017-02-03 | 2388.00 |
| 1778  | ABF    | 2017-02-04 | 2427.00 |
| 2044  | ABF    | 2017-02-05 | 2427.00 |
| 2309  | ABF    | 2017-02-06 | 2427.00 |
| 2575  | ABF    | 2017-02-07 | 2412.00 |
| 2859  | ABF    | 2017-02-08 | 2451.00 |
| 3159  | ABF    | 2017-02-09 | 2503.00 |
| 3475  | ABF    | 2017-02-10 | 2507.00 |
| 3797  | ABF    | 2017-02-11 | 2500.00 |
| 4125  | ABF    | 2017-02-12 | 2500.00 |
| 4453  | ABF    | 2017-02-13 | 2500.00 |
| 4779  | ABF    | 2017-02-14 | 2575.00 |
| 5111  | ABF    | 2017-02-15 | 2550.00 |
| 5450  | ABF    | 2017-02-16 | 2561.00 |
| 5798  | ABF    | 2017-02-17 | 2579.00 |
| 6151  | ABF    | 2017-02-18 | 2609.02 |
| 6507  | ABF    | 2017-02-19 | 2609.02 |
| 6863  | ABF    | 2017-02-20 | 2609.02 |
| 7219  | ABF    | 2017-02-21 | 2616.00 |
| 7580  | ABF    | 2017-02-22 | 2586.00 |
| 7951  | ABF    | 2017-02-23 | 2620.36 |
| 8635  | ABF    | 2017-02-24 | 2620.00 |
| 8963  | ABF    | 2017-02-25 | 2632.72 |
| 9291  | ABF    | 2017-02-28 | 2668.00 |
| 11551 | ABF    | 2017-03-08 | 2618.00 |
| 11842 | ABF    | 2017-03-09 | 2639.00 |
| 12190 | ABF    | 2017-03-10 | 2645.00 |
| 12538 | ABF    | 2017-03-13 | 2657.00 |
| 12894 | ABF    | 2017-03-14 | 2657.50 |

当前查询

SELECT 
ratings.id AS rating_id
MIN(price_date) AS first_correct,
DATEDIFF( MIN(price_date), issue_date ) AS days_lapsed
FROM
(
    SELECT
    ratings.id,
    ratings.ticker,
    broker,
    issue_date,
    close_price,
    price_date,
    ( ( ( price - close_price) / close_price ) * 100 ) AS diff_percent,
    FROM ratings
    JOIN prices
    ON ratings.ticker = prices.ticker
    WHERE price_date > issue_date
    HAVING diff_percent >= 5
) correct_ratings
GROUP BY ratings.id

结果

| rating_id | ticker | broker | issue_date | close_price | price_date | diff_percent | first_correct | days_lapsed |
|-----------|--------|--------|------------|-------------|------------|--------------|---------------|-------------|
| 22        | ABF    | HSBC   | 2017-01-23 | 2523        | 2017-02-28 | 5.747126     | 2017-02-28    | 36          |
| 775       | ABF    | HSBC   | 2017-02-15 | 2527        | 2017-02-28 | 5.579739     | 2017-02-28    | 13          |

问题

正如你所看到的first_correct日期为rating_id=22是后issue_date的rating_id=775 。 因此不应出现在表格中。 结果中只能rating_id=775 。

Answer 1

正如我的问题中提到的那样，问题在于我只有一个开始日期来检查百分比增加，而实际上我需要一个start_date 和 end_date 。

可以得出此结果的唯一方法是检查该经纪人/股票行情组合的下一个评级，并将其添加到名为end_date的新列中。

我能够使用在这里找到的查询来创建新列： https : //stackoverflow.com/a/15247545/1281101

有时，如果未取代该等级，则end_date将为NULL这可以通过IFNULL(column, return) MySQL函数来处理。

最终查询如下所示：

SELECT * 
FROM ( 
    SELECT rating_id, wp_broker_ratings.ticker, broker, date_issued AS start_date, issue_date_close_price, ( 
         SELECT MIN(st.date_issued) 
         FROM wp_broker_ratings st 
         WHERE st.broker = wp_broker_ratings.broker 
           AND st.ticker = wp_broker_ratings.ticker 
           AND st.date_issued > wp_broker_ratings.date_issued ) AS end_date, price_id, price_date, day_high AS absolute_correct_price, ( ( ( day_high - issue_date_close_price) / issue_date_close_price ) * 100 ) AS pct_diff 
FROM wp_broker_ratings 
INNER JOIN wp_share_prices 
ON wp_broker_ratings.ticker = wp_share_prices.ticker 
WHERE buy_sell = 'buy' 
HAVING price_date > start_date 
AND price_date <= IFNULL(end_date,'2050-01-01') 
AND pct_diff >= 5 ) correct 
GROUP BY rating_id

希望可以帮助别人。