寻求有关创建 sqlite 查询以根据 FIFO 方法处理买卖交易并生成合并摘要的建议
Seek advise on creating sqlite query to process buy sell transactions according to FIFO method and produce a combine summary
我正在尝试创建一个 sqlite 查询来处理基于 FIFO 方法的股票买卖交易，以跟踪计数器性能的摘要。
目前已写入 2 个查询：
- 撮合1买1卖交易，追查交易表现，详见下文“fifo_command_match_obos_query.txt”
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, qty, amount ,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s FROM trans_fifo WHERE txtype = 'sell'
)
SELECT
p.item Counter, round(d.amount - p.amount,2) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item and p.qty = d.qty and p.rn = d.s
WHERE PL NO NULL
;
生成此输出：
Counter PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
RSAWIT 173.3
SUCCESS 110.88
TNLOGIS 120.38
VELESTO 127.0
- 撮合多买一卖交易追踪交易表现，详见“fifo_command_match_mbos_query.txt”
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, qty, amount ,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s FROM trans_fifo WHERE txtype = 'sell'
)
SELECT
p.item Counter, round(d.amount - sum(p.amount),2) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item and p.qty <> d.qty
GROUP BY p.item
;
产生这个输出
Counter PL
------- ------
AEMULUS [NULL]
ARMADA [NULL]
ARREIT [NULL]
BAUTO [NULL]
CIMB [NULL]
D&O [NULL]
EFORCE [NULL]
GENM [NULL]
HEVEA [NULL]
IGBREIT [NULL]
IRIS [NULL]
JCY [NULL]
KPJ [NULL]
MBMR [NULL]
MBSB [NULL]
N2N 57.21
RSAWIT [NULL]
SUCCESS [NULL]
SUNREIT [NULL]
TNLOGIS [NULL]
TWREIT [NULL]
VELESTO [NULL]
VIS 331.69
YTLPOWR 88.19
我的请求是如何利用这些查询（或者请建议解决这种情况的更好方法）来编写有效的查询以生成输出以将两种情况都包含在一个摘要中，如下所示？
Counter PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
RSAWIT 173.3
SUCCESS 110.88
TNLOGIS 120.38
VELESTO 127.0
N2N 57.21
VIS 331.69
YTLPOWR 88.19
示例数据 trans_fifo table 用于下面的测试：
item date txtype qty price amount
------- ---------- ------ ----- ----- -------
AEMULUS 2020-09-17 buy 1000 0.67 678.62
AEMULUS 2020-10-23 sell 1000 0.735 726.36
ARMADA 2020-06-01 buy 2000 0.24 487.56
ARMADA 2020-09-02 sell 2000 0.3 591.4
ARREIT 2020-06-24 buy 2000 0.68 1370.95
CIMB 2021-01-21 buy 1000 3.98 3994.73
CIMB 2021-02-10 sell 1000 4.16 4145.21
D&O 2020-12-31 buy 1000 2.34 2352.24
D&O 2021-01-05 sell 1000 2.45 2437.73
EFORCE 2020-06-16 buy 2000 0.455 918.69
EFORCE 2020-07-28 sell 2000 0.51 1009.15
GENM 2020-02-26 buy 2000 2.93 5877.3
GENM 2020-12-07 buy 2000 2.56 5136.08
HEVEA 2020-09-11 buy 5000 0.435 2187.19
HEVEA 2020-09-29 sell 5000 0.48 2387.74
IGBREIT 2020-06-26 buy 1000 1.81 1822.08
IGBREIT 2021-01-08 buy 1000 1.68 1692.04
IGBREIT 2021-01-13 buy 1000 1.64 1652.03
IRIS 2020-07-03 buy 5000 0.2 1008.72
IRIS 2020-07-06 sell 5000 0.215 1064.14
JCY 2020-09-09 buy 2000 0.75 1511.99
JCY 2020-09-09 sell 2000 0.78 1547.99
JCY 2020-09-10 buy 2000 0.745 1500.99
JCY 2020-11-24 sell 2000 0.625 1239.09
KPJ 2020-10-22 buy 1800 0.88 1596.02
KPJ 2020-12-07 sell 1800 1.0 1787.92
MBMR 2020-10-30 buy 1000 2.63 2643.33
MBMR 2020-11-05 sell 1000 2.93 2916.58
MBMR 2020-11-10 buy 1000 2.9 2913.41
MBMR 2020-11-13 sell 1000 3.1 3086.53
MBSB 2021-01-13 buy 1000 0.625 633.61
MBSB 2021-02-16 sell 1000 0.665 656.38
N2N 2020-09-04 buy 2000 0.795 1602.02
N2N 2020-09-07 buy 1000 0.775 783.65
N2N 2020-09-10 buy 1000 0.695 703.63
N2N 2021-02-10 sell 4000 0.79 3146.51
RSAWIT 2020-09-24 buy 10000 0.26 2613.32
RSAWIT 2020-10-12 sell 10000 0.28 2786.62
SUCCESS 2020-07-27 buy 3000 0.56 1692.04
SUCCESS 2020-08-07 sell 3000 0.605 1802.92
SUCCESS 2021-01-22 buy 2000 0.79 1592.01
SUCCESS 2021-01-26 buy 2000 0.75 1511.99
SUCCESS 2021-02-18 buy 4000 0.74 2973.43
SUNREIT 2020-05-28 buy 1000 1.57 1582.01
SUNREIT 2020-10-06 buy 1000 1.5 1511.99
SUNREIT 2021-01-06 buy 1000 1.4 1410.96
TNLOGIS 2021-02-04 buy 5000 0.83 4164.79
TNLOGIS 2021-02-08 sell 5000 0.86 4285.17
TWREIT 2020-06-07 buy 2000 0.705 1420.96
VELESTO 2020-08-07 buy 10000 0.145 1460.98
VELESTO 2021-02-17 sell 10000 0.16 1587.98
VIS 2020-09-30 buy 5000 0.48 2412.26
VIS 2020-10-07 buy 1000 0.445 452.55
VIS 2020-12-07 sell 6000 0.535 3196.5
YTLPOWR 2020-02-07 buy 1000 0.725 733.64
YTLPOWR 2020-02-26 buy 2000 0.695 1400.96
YTLPOWR 2020-10-23 sell 3000 0.745 2222.79
BAUTO 2021-02-04 buy 5000 1.34 6718.55
BAUTO 2021-02-08 sell 5000 1.38 6881.39
trans_fifo 架构如下：
sqlite> .schema trans_fifo
CREATE VIEW trans_fifo
AS
SELECT
--broker.name as broker,
--investor.name as investor,
stock.counter as item,
trans.date as date,
type.tname as txtype,
trans.unit as qty,
trans.price as price,
trans.amount as amount
FROM
trans
INNER JOIN investor on investor.investor_id=trans.investor_id
INNER JOIN broker on broker.broker_id=trans.broker_id
INNER JOIN stock on stock.stock_id=trans.stock_id
INNER JOIN type on type.ttype=trans.ttype
WHERE
investor.name = 'SOON' AND broker.name = 'RAKUTEN'
-- stock.counter = 'SLP' AND investor.investor_id=1
ORDER BY
trans.broker_id ASC,
trans.stock_id ASC
/* trans_fifo(item,date,txtype,qty,price,amount) */;
感谢并感谢您花时间来遮蔽一些光线和见解
想了几天，下面是我自己贴的解决方法：
方案利用sqlitewindow框架函数添加caculation/manipulation变量达到目的
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn,
COUNT(item) OVER (PARTITION BY item) AS crn,
SUM(qty) OVER (PARTITION BY item ORDER By item, date ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING) AS bqsum,
SUM(amount) OVER (PARTITION BY item ORDER By item, date ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING) AS amsum
FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, date, qty, amount,
ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s
FROM trans_fifo WHERE txtype = 'sell' --GROUP BY item
)
SELECT
p.item,
(CASE
WHEN
p.item=out.item AND p.qty = d.qty AND p.rn = d.s THEN round(d.amount - p.amount,2)
WHEN
p.item=d.item AND p.bqsum = d.qty AND d.date >= p.date THEN round(d.amount - p.amsum,2)
END) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item
WHERE
PL IS NOT NULL
这是最终结果：
item PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
N2N 57.21
RSAWIT 173.3
SUCCESS 110.88
SUCCESS 265.12
TNLOGIS 120.38
VELESTO 127.0
VIS 331.69
YTLPOWR 88.19
我正在尝试创建一个 sqlite 查询来处理基于 FIFO 方法的股票买卖交易，以跟踪计数器性能的摘要。
目前已写入 2 个查询：
- 撮合1买1卖交易，追查交易表现，详见下文“fifo_command_match_obos_query.txt”
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, qty, amount ,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s FROM trans_fifo WHERE txtype = 'sell'
)
SELECT
p.item Counter, round(d.amount - p.amount,2) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item and p.qty = d.qty and p.rn = d.s
WHERE PL NO NULL
;
生成此输出：
Counter PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
RSAWIT 173.3
SUCCESS 110.88
TNLOGIS 120.38
VELESTO 127.0
- 撮合多买一卖交易追踪交易表现，详见“fifo_command_match_mbos_query.txt”
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, qty, amount ,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s FROM trans_fifo WHERE txtype = 'sell'
)
SELECT
p.item Counter, round(d.amount - sum(p.amount),2) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item and p.qty <> d.qty
GROUP BY p.item
;
产生这个输出
Counter PL
------- ------
AEMULUS [NULL]
ARMADA [NULL]
ARREIT [NULL]
BAUTO [NULL]
CIMB [NULL]
D&O [NULL]
EFORCE [NULL]
GENM [NULL]
HEVEA [NULL]
IGBREIT [NULL]
IRIS [NULL]
JCY [NULL]
KPJ [NULL]
MBMR [NULL]
MBSB [NULL]
N2N 57.21
RSAWIT [NULL]
SUCCESS [NULL]
SUNREIT [NULL]
TNLOGIS [NULL]
TWREIT [NULL]
VELESTO [NULL]
VIS 331.69
YTLPOWR 88.19
我的请求是如何利用这些查询（或者请建议解决这种情况的更好方法）来编写有效的查询以生成输出以将两种情况都包含在一个摘要中，如下所示？
Counter PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
RSAWIT 173.3
SUCCESS 110.88
TNLOGIS 120.38
VELESTO 127.0
N2N 57.21
VIS 331.69
YTLPOWR 88.19
示例数据 trans_fifo table 用于下面的测试：
item date txtype qty price amount
------- ---------- ------ ----- ----- -------
AEMULUS 2020-09-17 buy 1000 0.67 678.62
AEMULUS 2020-10-23 sell 1000 0.735 726.36
ARMADA 2020-06-01 buy 2000 0.24 487.56
ARMADA 2020-09-02 sell 2000 0.3 591.4
ARREIT 2020-06-24 buy 2000 0.68 1370.95
CIMB 2021-01-21 buy 1000 3.98 3994.73
CIMB 2021-02-10 sell 1000 4.16 4145.21
D&O 2020-12-31 buy 1000 2.34 2352.24
D&O 2021-01-05 sell 1000 2.45 2437.73
EFORCE 2020-06-16 buy 2000 0.455 918.69
EFORCE 2020-07-28 sell 2000 0.51 1009.15
GENM 2020-02-26 buy 2000 2.93 5877.3
GENM 2020-12-07 buy 2000 2.56 5136.08
HEVEA 2020-09-11 buy 5000 0.435 2187.19
HEVEA 2020-09-29 sell 5000 0.48 2387.74
IGBREIT 2020-06-26 buy 1000 1.81 1822.08
IGBREIT 2021-01-08 buy 1000 1.68 1692.04
IGBREIT 2021-01-13 buy 1000 1.64 1652.03
IRIS 2020-07-03 buy 5000 0.2 1008.72
IRIS 2020-07-06 sell 5000 0.215 1064.14
JCY 2020-09-09 buy 2000 0.75 1511.99
JCY 2020-09-09 sell 2000 0.78 1547.99
JCY 2020-09-10 buy 2000 0.745 1500.99
JCY 2020-11-24 sell 2000 0.625 1239.09
KPJ 2020-10-22 buy 1800 0.88 1596.02
KPJ 2020-12-07 sell 1800 1.0 1787.92
MBMR 2020-10-30 buy 1000 2.63 2643.33
MBMR 2020-11-05 sell 1000 2.93 2916.58
MBMR 2020-11-10 buy 1000 2.9 2913.41
MBMR 2020-11-13 sell 1000 3.1 3086.53
MBSB 2021-01-13 buy 1000 0.625 633.61
MBSB 2021-02-16 sell 1000 0.665 656.38
N2N 2020-09-04 buy 2000 0.795 1602.02
N2N 2020-09-07 buy 1000 0.775 783.65
N2N 2020-09-10 buy 1000 0.695 703.63
N2N 2021-02-10 sell 4000 0.79 3146.51
RSAWIT 2020-09-24 buy 10000 0.26 2613.32
RSAWIT 2020-10-12 sell 10000 0.28 2786.62
SUCCESS 2020-07-27 buy 3000 0.56 1692.04
SUCCESS 2020-08-07 sell 3000 0.605 1802.92
SUCCESS 2021-01-22 buy 2000 0.79 1592.01
SUCCESS 2021-01-26 buy 2000 0.75 1511.99
SUCCESS 2021-02-18 buy 4000 0.74 2973.43
SUNREIT 2020-05-28 buy 1000 1.57 1582.01
SUNREIT 2020-10-06 buy 1000 1.5 1511.99
SUNREIT 2021-01-06 buy 1000 1.4 1410.96
TNLOGIS 2021-02-04 buy 5000 0.83 4164.79
TNLOGIS 2021-02-08 sell 5000 0.86 4285.17
TWREIT 2020-06-07 buy 2000 0.705 1420.96
VELESTO 2020-08-07 buy 10000 0.145 1460.98
VELESTO 2021-02-17 sell 10000 0.16 1587.98
VIS 2020-09-30 buy 5000 0.48 2412.26
VIS 2020-10-07 buy 1000 0.445 452.55
VIS 2020-12-07 sell 6000 0.535 3196.5
YTLPOWR 2020-02-07 buy 1000 0.725 733.64
YTLPOWR 2020-02-26 buy 2000 0.695 1400.96
YTLPOWR 2020-10-23 sell 3000 0.745 2222.79
BAUTO 2021-02-04 buy 5000 1.34 6718.55
BAUTO 2021-02-08 sell 5000 1.38 6881.39
trans_fifo 架构如下：
sqlite> .schema trans_fifo
CREATE VIEW trans_fifo
AS
SELECT
--broker.name as broker,
--investor.name as investor,
stock.counter as item,
trans.date as date,
type.tname as txtype,
trans.unit as qty,
trans.price as price,
trans.amount as amount
FROM
trans
INNER JOIN investor on investor.investor_id=trans.investor_id
INNER JOIN broker on broker.broker_id=trans.broker_id
INNER JOIN stock on stock.stock_id=trans.stock_id
INNER JOIN type on type.ttype=trans.ttype
WHERE
investor.name = 'SOON' AND broker.name = 'RAKUTEN'
-- stock.counter = 'SLP' AND investor.investor_id=1
ORDER BY
trans.broker_id ASC,
trans.stock_id ASC
/* trans_fifo(item,date,txtype,qty,price,amount) */;
感谢并感谢您花时间来遮蔽一些光线和见解
想了几天，下面是我自己贴的解决方法：
方案利用sqlitewindow框架函数添加caculation/manipulation变量达到目的
WITH Purchase AS (
SELECT *,ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS rn,
COUNT(item) OVER (PARTITION BY item) AS crn,
SUM(qty) OVER (PARTITION BY item ORDER By item, date ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING) AS bqsum,
SUM(amount) OVER (PARTITION BY item ORDER By item, date ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING) AS amsum
FROM trans_fifo WHERE txtype = 'buy'
),
Dispose AS (
SELECT item, date, qty, amount,
ROW_NUMBER() OVER (PARTITION BY item ORDER BY date) AS s
FROM trans_fifo WHERE txtype = 'sell' --GROUP BY item
)
SELECT
p.item,
(CASE
WHEN
p.item=out.item AND p.qty = d.qty AND p.rn = d.s THEN round(d.amount - p.amount,2)
WHEN
p.item=d.item AND p.bqsum = d.qty AND d.date >= p.date THEN round(d.amount - p.amsum,2)
END) PL
FROM
Purchase p
LEFT JOIN
Dispose d
ON
p.item = d.item
WHERE
PL IS NOT NULL
这是最终结果：
item PL
------- ------
AEMULUS 47.74
ARMADA 103.84
BAUTO 162.84
CIMB 150.48
D&O 85.49
EFORCE 90.46
HEVEA 200.55
IRIS 55.42
JCY 36.0
JCY -261.9
KPJ 191.9
MBMR 273.25
MBMR 173.12
MBSB 22.77
N2N 57.21
RSAWIT 173.3
SUCCESS 110.88
SUCCESS 265.12
TNLOGIS 120.38
VELESTO 127.0
VIS 331.69
YTLPOWR 88.19