How to calculate percentage increase from previous row/day after complex Group By?

问题

I have a table IntradayPrices1Minute where I store 1 minute timeframe open, high, low and close prices for stocks:

  CREATE TABLE `IntradayPrices1Minute` (
  `ticker` varchar(10) NOT NULL DEFAULT '',
  `datetime` datetime NOT NULL,
  `volume` mediumint(11) unsigned NOT NULL,
  `open` decimal(8,4) unsigned NOT NULL,
  `high` decimal(8,4) unsigned NOT NULL,
  `low` decimal(8,4) unsigned NOT NULL,
  `close` decimal(8,4) unsigned NOT NULL,
  PRIMARY KEY (`datetime`,`ticker`),
  UNIQUE KEY `indxTickerDatetime` (`ticker`,`datetime`) USING BTREE
)

I have successfully build a query where I can calculate the daily open, high, low and close prices for those stocks. This is the query:

SELECT 
    ticker,
    DATE(datetime) AS 'Date',
    SUBSTRING_INDEX( GROUP_CONCAT(CAST(open AS CHAR) ORDER BY datetime), ',', 1 ) as 'Daily Open',
    max(GREATEST(open, high, low, close)) AS 'Daily High',
    min(LEAST(open, high, low, close)) AS 'Daily Low',
    SUBSTRING_INDEX( GROUP_CONCAT(CAST(close AS CHAR) ORDER BY datetime DESC), ',', 1 ) as 'Daily Close'

FROM 
    IntradayPrices1Minute

GROUP BY
    ticker, date(datetime)

and this is part of the results that this query successfully returns:

ticker  Date        Open    High    Low     Close
----    ----------  ------  ------  ------  ------ 
AAAE    2012-11-26  0.0100  0.0100  0.0100  0.0100
AAAE    2012-11-27  0.0130  0.0140  0.0083  0.0140
AAAE    2012-11-28  0.0140  0.0175  0.0140  0.0165
AAAE    2012-11-29  0.0175  0.0175  0.0137  0.0137
AAMRQ   2012-11-26  0.4411  0.5300  0.4411  0.5290
AAMRQ   2012-11-27  0.5100  0.5110  0.4610  0.4950
AAMRQ   2012-11-28  0.4820  0.4900  0.4300  0.4640
AAMRQ   2012-11-29  0.4505  0.4590  0.4411  0.4590
AAMRQ   2012-11-30  0.4500  0.4570  0.4455  0.4568

Now the problem is: I want to return a seventh column in the query that calculates for each day the percentage increase/decrease between its close price and the previous day close price.

I have seen similar questions asked on StackOverflow but for situations in which the daily prices are already in a table. I think it is specially complex in my case because the daily prices are obtained at query time after several grouping calculations.

Any help on this would be greatly appreciated. Many Thanks. Boga

回答1:

Please take a look at this reference: SQLFIDDLE

So your current table with OHLC is used to derive the px_change, px_pct :) Results have been rounded to show 4 decimal points in change as well as percentage. Null is used to show no price change/pct as null is a better representation of no price change than having a zero price change ;)

Query:

select ticker, date_format(date,'%m-%d-%Y') as date, open, high,low,close,
pxchange,concat(round(pxpct*100,4),'%') pxpct
from (select case when ticker <> @pxticker 
        then @pxclose := null end, p.*, (close-@pxclose) as pxchange,
    (close-@pxclose)/@pxclose as pxpct, (@pxclose := close),
    (@pxticker := ticker) from pricing p
cross join
    (select @pxclose := null, @pxticker := ticker
     from pricing 
     order by ticker, date limit 1)  as a
  order by ticker, date ) as b
order by ticker, date asc

Resutls:

TICKER  DATE        OPEN    HIGH    LOW     CLOSE   PXCHANGE  PXPCT
AAAE    11-26-2012  0.01    0.01    0.01    0.01    (null)    (null)
AAAE    11-27-2012  0.013   0.014   0.0083  0.014   0.004     40.0000%
AAAE    11-28-2012  0.014   0.0175  0.014   0.0165  0.0025    17.8571%
AAAE    11-29-2012  0.0175  0.0175  0.0137  0.0137  -0.0028   -16.9697%
AAMRQ   11-26-2012  0.4411  0.53    0.4411  0.529   (null)    (null)
AAMRQ   11-27-2012  0.51    0.511   0.461   0.495   -0.034    -6.4272%
AAMRQ   11-28-2012  0.482   0.49    0.43    0.464   -0.031    -6.2626%
AAMRQ   11-29-2012  0.4505  0.459   0.4411  0.459   -0.005    -1.0776%
AAMRQ   11-30-2012  0.45    0.457   0.4455  0.4568  -0.0022   -0.4793%

** UPDATED WITH A PICTURE TO SHOW THE PARENTHESIS AS PER OP'S REQUEST IN COMMENTS ** :)

回答2:

high should already contain the greatest and low should already contain the least value.

First, I'd rewrite your select and create it as a view as

create view daily_prices as
SELECT 
    ticker,
    DATE(datetime) AS 'Date',
    SUBSTRING_INDEX( GROUP_CONCAT(CAST(open AS CHAR) ORDER BY datetime), ',', 1 ) as 'Daily Open',
    max(high) AS 'Daily High',
    min(low) AS 'Daily Low',
    SUBSTRING_INDEX( GROUP_CONCAT(CAST(close AS CHAR) ORDER BY datetime DESC), ',', 1 ) as 'Daily Close'
FROM 
    IntradayPrices1Minute
GROUP BY
    ticker, date(datetime)

then you can do a self join

select d1.*, d1.`daily close` / d2.`daily close` - 1 as `percent change`
from daily_prices d1
join daily_prices d2 on d1.ticker = d2.ticker and d1.`date` = adddate(d2.`date`, 1)

SQL Fiddle

来源：https://stackoverflow.com/questions/13671230/how-to-calculate-percentage-increase-from-previous-row-day-after-complex-group-b

标签

mysql

group-by