无法在 SQL 中将日期放在列中，将指标放在另一个列中

Question

问题：对于给定的第 t 天，我需要为卖家 ID“123”创建一个 61 天的准时交货 (OTD) 指标。 OTD 是按时交付包裹的比率 对于在 t 之前 30 天内创建的包。
我们不仅想知道今天的OTD，而且我们想知道OTD 过去 60 天内的每个日期。

CREATE TABLE `packages` (
  `id` int NOT NULL,
  `created_at` timestamp NULL DEFAULT NULL,
  `seller_id` int DEFAULT NULL,
  `promise_date` date DEFAULT NULL,
  `delivered_at` timestamp NULL DEFAULT NULL,
  PRIMARY KEY (`id`)
);
INSERT INTO `packages` (`id`, `created_at`, `seller_id`, `promise_date`, `delivered_at`) VALUES
('1', '2020-04-15 06:26:28', '246', '2020-04-20', '2020-04-20 07:47:07'),
('2', '2020-05-02 06:19:13', '123', '2020-05-06', '2020-05-07 07:01:26'),
('3', '2020-07-05 08:47:17', '789', '2020-07-08', '2020-07-09 08:22:08'),
('4', '2020-03-25 08:12:35', '234', '2020-03-30', '2020-03-30 08:49:50'),
('5', '2020-07-01 07:51:33', '789', '2020-07-06', '2020-07-06 08:34:48'),
('6', '2020-03-01 08:11:31', '123', '2020-03-04', '2020-03-04 06:00:14'),
('7', '2020-05-20 07:38:14', '246', '2020-05-25', '2020-05-25 08:13:51'),
('8', '2020-04-14 07:30:19', '123', '2020-04-17', '2020-04-17 07:46:55'),
('9', '2020-02-10 08:53:00', '234', '2020-02-13', '2020-02-14 06:45:57'),
('10', '2020-01-21 07:14:56', '246', '2020-01-24', '2020-01-25 08:03:04'),
('11', '2020-07-03 08:05:11', '123', '2020-07-08', '2020-07-09 08:30:56'),
('12', '2020-05-09 06:18:31', '789', '2020-05-13', '2020-05-13 08:38:55'),
('13', '2020-02-13 08:11:10', '123', '2020-02-18', '2020-02-18 07:48:52'),
('14', '2020-04-28 08:25:28', '789', '2020-05-01', '2020-05-02 06:06:32'),
('15', '2020-06-02 07:28:52', '234', '2020-06-05', '2020-06-06 07:29:43'),
('16', '2020-05-04 08:39:33', '123', '2020-05-07', '2020-05-08 06:33:14'),
('17', '2020-07-26 08:18:30', '789', '2020-07-29', '2020-07-30 07:28:53'),
('18', '2020-02-25 08:37:42', '234', '2020-02-28', '2020-02-29 06:05:23'),
('19', '2020-02-03 06:55:39', '234', '2020-02-06', '2020-02-07 07:18:28'),
('20', '2020-03-07 08:20:44', '246', '2020-03-11', '2020-03-11 08:11:45'),
('21', '2020-03-11 07:19:47', '789', '2020-03-16', '2020-03-16 06:55:46'),
('22', '2020-06-24 08:18:56', '789', '2020-06-29', '2020-06-29 08:47:59'),
('23', '2020-02-25 07:24:19', '123', '2020-02-28', '2020-02-28 06:54:57'),
('24', '2020-07-12 07:51:52', '789', '2020-07-15', '2020-07-16 07:36:21'),
('25', '2020-01-26 07:44:59', '234', '2020-01-29', '2020-01-29 08:52:24'),
('26', '2020-02-07 06:09:24', '246', '2020-02-12', '2020-02-13 08:16:37'),
('27', '2020-03-11 08:34:57', '123', '2020-03-16', '2020-03-17 08:33:47'),
('28', '2020-02-24 08:15:41', '789', '2020-02-27', '2020-02-27 06:19:59'),
('29', '2020-02-02 06:45:36', '123', '2020-02-05', '2020-02-06 06:22:25'),
('30', '2020-02-10 06:51:48', '123', '2020-02-13', '2020-02-13 06:45:07'),
('31', '2020-03-27 07:11:58', '789', '2020-04-01', '2020-04-02 08:55:56'),
('32', '2020-05-31 07:10:05', '246', '2020-06-03', '2020-06-03 08:56:47'),
('33', '2020-06-28 06:14:19', '789', '2020-07-01', '2020-07-02 06:35:18'),
('34', '2020-07-08 08:30:12', '789', '2020-07-13', '2020-07-14 06:06:09'),
('35', '2020-05-13 08:13:34', '123', '2020-05-18', '2020-05-18 08:24:42'),
('36', '2020-04-19 08:13:38', '246', '2020-04-22', '2020-04-22 07:32:14'),
('37', '2020-03-02 06:57:32', '234', '2020-03-05', '2020-03-05 07:16:05'),
('38', '2020-05-22 08:49:51', '246', '2020-05-27', '2020-05-27 06:47:41'),
('39', '2020-02-27 08:18:26', '123', '2020-03-03', '2020-03-03 06:32:56'),
('40', '2020-02-17 07:10:24', '246', '2020-02-20', '2020-02-21 06:06:26'),
('41', '2020-06-25 08:29:32', '234', '2020-06-30', '2020-06-30 07:37:07'),
('42', '2020-03-02 08:07:57', '234', '2020-03-05', '2020-03-05 08:41:13'),
('43', '2020-06-18 06:44:38', '234', '2020-06-23', '2020-06-23 06:11:26'),
('44', '2020-07-15 08:22:49', '246', '2020-07-20', '2020-07-20 08:34:28'),
('45', '2020-07-07 07:54:10', '123', '2020-07-10', '2020-07-10 07:50:24'),
('46', '2020-07-17 07:43:08', '123', '2020-07-22', '2020-07-22 06:33:22'),
('47', '2020-04-01 08:24:20', '234', '2020-04-06', '2020-04-06 06:12:55'),
('48', '2020-05-14 08:49:10', '123', '2020-05-19', '2020-05-20 06:53:50'),
('49', '2020-06-11 08:20:35', '246', '2020-06-16', '2020-06-16 06:21:10'),
('50', '2020-06-24 06:39:29', '789', '2020-06-29', '2020-06-30 06:03:48'),
('51', '2020-02-29 06:43:01', '246', '2020-03-04', '2020-03-05 07:57:51'),
('52', '2020-07-17 08:23:46', '246', '2020-07-22', '2020-07-22 07:49:01'),
('53', '2020-03-27 07:45:10', '123', '2020-04-01', '2020-04-02 06:43:34'),
('54', '2020-04-28 06:39:55', '246', '2020-05-01', '2020-05-01 08:59:08'),
('55', '2020-05-21 07:16:03', '789', '2020-05-26', '2020-05-26 06:29:53'),
('56', '2020-02-10 08:22:04', '246', '2020-02-13', '2020-02-14 06:24:17'),
('57', '2020-02-02 08:04:26', '234', '2020-02-05', '2020-02-05 08:59:43'),
('58', '2020-03-02 08:21:53', '246', '2020-03-05', '2020-03-06 08:45:36'),
('59', '2020-02-19 08:37:15', '123', '2020-02-24', '2020-02-24 08:24:44'),
('60', '2020-06-16 08:51:24', '234', '2020-06-19', '2020-06-20 08:25:14'),
('61', '2020-07-11 07:37:15', '234', '2020-07-15', '2020-07-15 07:03:13'),
('62', '2020-06-15 07:56:39', '123', '2020-06-18', '2020-06-19 07:11:16'),
('63', '2020-03-06 07:21:52', '123', '2020-03-11', '2020-03-11 07:46:48'),
('64', '2020-06-03 06:43:50', '789', '2020-06-08', '2020-06-09 07:40:17'),
('65', '2020-01-20 06:28:47', '234', '2020-01-23', '2020-01-24 08:34:05'),
('66', '2020-04-02 08:04:41', '123', '2020-04-07', '2020-04-08 08:56:45'),
('67', '2020-03-04 06:05:57', '789', '2020-03-09', '2020-03-10 06:26:56'),
('68', '2020-07-04 06:47:46', '246', '2020-07-08', '2020-07-09 06:53:02'),
('69', '2020-02-25 06:47:09', '246', '2020-02-28', '2020-02-28 07:55:25'),
('70', '2020-02-04 07:17:28', '123', '2020-02-07', '2020-02-07 08:07:54'),
('71', '2020-06-15 07:18:16', '789', '2020-06-18', '2020-06-19 06:02:08'),
('72', '2020-07-09 06:32:34', '234', '2020-07-14', '2020-07-14 08:15:02'),
('73', '2020-05-21 06:12:52', '789', '2020-05-26', '2020-05-27 07:39:20'),
('74', '2020-05-24 06:38:49', '789', '2020-05-27', '2020-05-27 06:51:35'),
('75', '2020-02-27 06:31:02', '123', '2020-03-03', '2020-03-03 08:56:26'),
('76', '2020-07-02 08:55:00', '123', '2020-07-07', '2020-07-07 07:42:16'),
('77', '2020-06-30 06:52:27', '246', '2020-07-03', '2020-07-03 07:43:20'),
('78', '2020-04-25 08:08:14', '246', '2020-04-29', '2020-04-29 07:21:23'),
('79', '2020-06-24 08:34:43', '234', '2020-06-29', '2020-06-30 06:43:59'),
('80', '2020-05-13 08:59:11', '246', '2020-05-18', '2020-05-18 07:19:06'),
('81', '2020-02-21 07:14:16', '789', '2020-02-26', '2020-02-27 07:10:39'),
('82', '2020-06-04 08:43:13', '789', '2020-06-09', '2020-06-09 07:24:28'),
('83', '2020-07-04 07:14:42', '234', '2020-07-08', '2020-07-09 07:45:59'),
('84', '2020-05-24 08:17:00', '246', '2020-05-27', '2020-05-27 06:31:15'),
('85', '2020-03-07 07:43:27', '123', '2020-03-11', '2020-03-12 08:39:45');



create table dates(
fulldate date);

INSERT INTO `dates` (`fulldate`) VALUES
('2020-08-01'),
('2020-07-31'),
('2020-07-30'),
('2020-07-29'),
('2020-07-28'),
('2020-07-27'),
('2020-07-26'),
('2020-07-25'),
('2020-07-24'),
('2020-07-23'),
('2020-07-22'),
('2020-07-21'),
('2020-07-20'),
('2020-07-19'),
('2020-07-18'),
('2020-07-17'),
('2020-07-16'),
('2020-07-15'),
('2020-07-14'),
('2020-07-13'),
('2020-07-12'),
('2020-07-11'),
('2020-07-10'),
('2020-07-09'),
('2020-07-08'),
('2020-07-07'),
('2020-07-06'),
('2020-07-05'),
('2020-07-04'),
('2020-07-03'),
('2020-07-02'),
('2020-07-01'),
('2020-06-30'),
('2020-06-29'),
('2020-06-28'),
('2020-06-27'),
('2020-06-26'),
('2020-06-25'),
('2020-06-24'),
('2020-06-23'),
('2020-06-22'),
('2020-06-21'),
('2020-06-20'),
('2020-06-19'),
('2020-06-18'),
('2020-06-17'),
('2020-06-16'),
('2020-06-15'),
('2020-06-14'),
('2020-06-13'),
('2020-06-12'),
('2020-06-11'),
('2020-06-10'),
('2020-06-09'),
('2020-06-08'),
('2020-06-07'),
('2020-06-06'),
('2020-06-05'),
('2020-06-04'),
('2020-06-03'),
('2020-06-02');

所需的输出：

date       | on_time_delivery_ratio

2020-07-31 | 0.75  
2020-07-30 | 0.69  
2020-07-29 | 0.68  
2020-07-28 | 0.80  
2020-07-27 | 0.79  
2020-07-26 | 0.78  
2020-07-25 | 0.69  
2020-07-24 | 0.72

我做了什么：
我已经能够创建一个指标，但只是针对当前日期。

SELECT curdate(),sum(case when promise_date=date(delivered_at) then 1 else 0 end)/count(*)*100 
 "On-Time Delivery Rate (%)"
from packages p 
where p.seller_id=123 and date(created_at) between DATE_SUB(curdate(),interval 30 day) and 
DATE_SUB(curdate(),interval 1 day); 

我需要帮助的地方：
我需要日期表中的日期，而不是使用 curdate()，作为所需输出的第一列和这些日期的指标。

SQL 小提琴： http://sqlfiddle.com/#!9/b665ca1/2

Answer 1

小提琴：http://sqlfiddle.com/#!9/b665ca1/7（基于http://sqlfiddle.com/#!9/b665ca1/2）

SELECT 
   delivered_at,
   sum(case when promise_date=date(delivered_at) then 1 else 0 end)/count(*)  "On-Time Delivery Rate (%)"
from packages p 
where p.seller_id=123 
  and date(delivered_at) between DATE_SUB(curdate(),interval 30 day) and DATE_SUB(curdate(),interval 1 day)
group by delivered_at;

输出：

delivered_at    On-Time Delivery Rate (%)
2020-07-07T07:42:16Z    1
2020-07-09T08:30:56Z    0
2020-07-10T07:50:24Z    1
2020-07-22T06:33:22Z    1

编辑：选择过去 61 天

SELECT date.d,sum(case when promise_date=date(delivered_at) then 1 else 0 end)/count(*)*100 
 "On-Time Delivery Rate (%)"
from date 
CROSS JOIN packages p 
where p.seller_id=123 
  and date(created_at) between DATE_SUB(date.d,interval 30 day) and DATE_SUB(date.d,interval 1 day)
  and date.d >= DATE_SUB(curdate(),interval 61 day)
GROUP BY date.d
ORDER BY date.d; 

• 我将您查询中的 curdate() 更改为值 date.d。
• select * from date 应提供所有日期，至少从过去 61 天开始。

Answer 2

您想在dates table 中获取所有日期的 OTD 百分比，并获取相应日期的指标。 因此百分比计算为：（（过去 30 天和日期前一天之间的准时交货次数）/（过去 30 天和日期前一天之间的总交货次数））*100；

因此，如果您以日期 2020-06-02 为例，则在 2020-05-02 至 2020-06-01 之间共有 3 次交付，其中 1 次按时交付，因此 OTD 百分比 = 33.3

据我了解，这是您可能正在寻找的解决方案：

请查看SQL fiddle:

SELECT d.fulldate,sum(case when promise_date = date(delivered_at) 
                      and p.seller_id = 123 and date(created_at) between DATE_SUB(d.fulldate,interval 30 day) and 
DATE_SUB(d.fulldate,interval 1 day) then 1 else 0 end)/sum(case when date(created_at) between DATE_SUB(d.fulldate,interval 30 day) and 
DATE_SUB(d.fulldate,interval 1 day) and p.seller_id = 123 then 1 else 0 end)*100
"On-Time Delivery Rate (%)"
from packages p, dates d group by d.fulldate;
where p.seller_id=123 and date(created_at) between DATE_SUB(d.fulldate,interval 30 day) and 
DATE_SUB(d.fulldate,interval 1 day);

输出：

以下是部分输出：（对于完整输出，您可以查看查询上方给出的 Fiddle）

fulldate    On-Time Delivery Rate (%)
2020-06-02  33.3333
2020-06-03  33.3333
2020-06-04  50
2020-06-05  50
2020-06-06  50
2020-06-07  50