DATEFLOAT
Updated 2023-10-06 23:39:48.637000
Syntax
SELECT [westclintech].[wct].[DATEFLOAT](
<@Year, int,>
,<@Month, int,>
,<@Day, int,>)
Description
Use the scalar functions DATEFLOAT to return a float value for a specified Year, Month, and Day. DATEFLOAT is the equivalent of the DATE function in EXCEL.
Arguments
@Year
The integer value representing the year. @Year is an expression that returns an int or that can be implicitly converted to int.
@Month
The integer value representing the month. @Month is an expression that returns an int or that can be implicitly converted to int.
@Day
The integer value representing the day of the month. @Day is an expression that returns an int or that can be implicitly converted to int.
Return Type
float
Remarks
DATEFLOAT is the equivalent of the DATE function in EXCEL, except that it returns a float representing the number of days from 01-Jan-1900. This is consistent with numeric representation of dates in SQL Server. EXCEL and .NET have different numeric representations of dates than SQL Server.
To convert a year, month, and day combination to a datetime, consider using the CALCDATE function.
To convert a year, month, and day combination to an integer, consider using the DATEINT function.
Examples
To represent 07-Feb-2011, enter the following SQL.
SELECT wct.DATEFLOAT( 2011, --@Year
2, --@Month
7 --@Day
) as DATEFLOAT;
This produces the following result.
{"columns":[{"field":"DATEFLOAT","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"DATEFLOAT":"40579"}]}
DATEFLOAT is language independent. The datetime value returned is consistent across different language settings. Look at the following example.
SET LANGUAGE Italian;
SELECT wct.DATEFLOAT(2011, 2, 7) as [DATEFLOAT(2011,2,7)],
cast(cast('02/07/2011' as datetime) as int) as [cast(cast('02/07/2011' as
datetime) as int)];
This returns the following result.
{"columns":[{"field":"DATEFLOAT(2011,2,7)","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"},{"field":"cast(cast('02/07/2011' as datetime) as int)","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"DATEFLOAT(2011,2,7)":"40579","cast(cast('02/07/2011' as datetime) as int)":"40724"}]}
SET LANGUAGE us_english;
SELECT wct.DATEFLOAT(2011, 2, 7) as [DATEFLOAT(2011,2,7)],
cast(cast('02/07/2011' as datetime) as int) as [cast(cast('02/07/2011' as
datetime) as int)];
This returns the following result.
{"columns":[{"field":"DATEFLOAT(2011,2,7)","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"},{"field":"cast(cast('02/07/2011' as datetime) as int)","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"DATEFLOAT(2011,2,7)":"40579","cast(cast('02/07/2011' as datetime) as int)":"40579"}]}
As you can see ,the same date string was passed in both statements, but the datetime value returned was different based upon the language setting for the database. Using DATEFLOAT avoids that problem.
DATEFLOAT allows you to perform operation on the date components. You can add or subtract integer values to @Year, @Month, @Date, and still be confident of having a valid value returned.
Let’s say you wanted to generate a schedule that showed the first Monday of every month. Here’s one way to do this in SQL.
SELECT DATEADD(
WEEK,
DATEDIFF(
WEEK,
0,
(DATEADD(
DAY,
(6 - DATEPART(DAY, DATEFROMPARTS(2011,
m.mth, 1))),
DATEFROMPARTS(2011, m.mth, 1)
)
)
),
0
)
FROM
(
VALUES
(1),
(2),
(3),
(4),
(5),
(6),
(7),
(8),
(9),
(10),
(11),
(12)
) m (mth);
This produces the following result.
{"columns":[{"field":"First Monday","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"First Monday":"2011-01-03 00:00:00.000"},{"First Monday":"2011-02-07 00:00:00.000"},{"First Monday":"2011-03-07 00:00:00.000"},{"First Monday":"2011-04-04 00:00:00.000"},{"First Monday":"2011-05-02 00:00:00.000"},{"First Monday":"2011-06-06 00:00:00.000"},{"First Monday":"2011-07-04 00:00:00.000"},{"First Monday":"2011-08-01 00:00:00.000"},{"First Monday":"2011-09-05 00:00:00.000"},{"First Monday":"2011-10-03 00:00:00.000"},{"First Monday":"2011-11-07 00:00:00.000"},{"First Monday":"2011-12-05 00:00:00.000"}]}
DATEFLOAT allows you to perform operations on the components of the date rather than on the date value, simplifying the SQL.
SELECT CAST(wct.DATEFLOAT(2011
,mth
,1 +(9 - DATEPART(DW,wct.DATEFLOAT(2011,mth,1))) % 7
) as datetime) as [First Monday]
FROM (
VALUES
(1),(2),(3),(4),(5),(6),
(7),(8),(9),(10),(11),(12)
) m(mth);
This produces the following result.
{"columns":[{"field":"First Monday","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"First Monday":"2011-01-03 00:00:00.000"},{"First Monday":"2011-02-07 00:00:00.000"},{"First Monday":"2011-03-07 00:00:00.000"},{"First Monday":"2011-04-04 00:00:00.000"},{"First Monday":"2011-05-02 00:00:00.000"},{"First Monday":"2011-06-06 00:00:00.000"},{"First Monday":"2011-07-04 00:00:00.000"},{"First Monday":"2011-08-01 00:00:00.000"},{"First Monday":"2011-09-05 00:00:00.000"},{"First Monday":"2011-10-03 00:00:00.000"},{"First Monday":"2011-11-07 00:00:00.000"},{"First Monday":"2011-12-05 00:00:00.000"}]}
As a side note, the reason that we used 9 to represent Monday and did not use 2 in this SQL (2 being the numeric representation of Monday), is that the modulo function, %, in SQL Server is really a remainder function. The SQL would look more intuitive if we had used the XLeratorDB modulo function (which is the equivalent of the EXCEL MOD function).
SELECT CAST(fom + wct.Modulo(-wct.Modulo(fom, 7), 7) as datetime) as [First Monday]
FROM
(
SELECT wct.DATEFLOAT(2011, mth, 1) as fom
FROM
(
VALUES
(1),
(2),
(3),
(4),
(5),
(6),
(7),
(8),
(9),
(10),
(11),
(12)
) m (mth)
) n;
This produces the following result.
{"columns":[{"field":"First Monday","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"First Monday":"2011-01-03 00:00:00.000"},{"First Monday":"2011-02-07 00:00:00.000"},{"First Monday":"2011-03-07 00:00:00.000"},{"First Monday":"2011-04-04 00:00:00.000"},{"First Monday":"2011-05-02 00:00:00.000"},{"First Monday":"2011-06-06 00:00:00.000"},{"First Monday":"2011-07-04 00:00:00.000"},{"First Monday":"2011-08-01 00:00:00.000"},{"First Monday":"2011-09-05 00:00:00.000"},{"First Monday":"2011-10-03 00:00:00.000"},{"First Monday":"2011-11-07 00:00:00.000"},{"First Monday":"2011-12-05 00:00:00.000"}]}
DATEFLOAT can simplify date calculations on the database. Let’s say that you are a company that has a 52-53 week fiscal year, and the year end is defined as the final Saturday in the year end month. You need to calculate the year end date and the number of days in the year.
DECLARE @YEM as int = 8;
SELECT yr,
CAST(ye - wct.Modulo(wct.Modulo(ye, 7) - 5, 7) as datetime) as [Last Saturday]
,
(ye - wct.Modulo(wct.Modulo(ye, 7) - 5, 7)) - (pye - wct.Modulo(wct.Modulo(
pye, 7) - 5, 7)) as [Days In Year]
FROM
(
SELECT yr,
wct.DATEFLOAT(n.yr, @yem + 1, 1) - 1 as ye,
wct.DATEFLOAT(n.yr - 1, @yem + 1, 1) - 1 as pye
FROM
(
VALUES
(2005),
(2006),
(2007),
(2008),
(2009),
(2010),
(2011),
(2012),
(2013),
(2014),
(2015),
(2016),
(2017),
(2018),
(2019)
) n (yr)
) m;
This produces the following result.
{"columns":[{"field":"yr","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"},{"field":"Last Saturday","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"},{"field":"Days In Year","headerClass":"ag-right-aligned-header","cellClass":"ag-right-aligned-cell"}],"rows":[{"yr":"2005","Last Saturday":"2005-08-27 00:00:00.000","Days In Year":"364"},{"yr":"2006","Last Saturday":"2006-08-26 00:00:00.000","Days In Year":"364"},{"yr":"2007","Last Saturday":"2007-08-25 00:00:00.000","Days In Year":"364"},{"yr":"2008","Last Saturday":"2008-08-30 00:00:00.000","Days In Year":"371"},{"yr":"2009","Last Saturday":"2009-08-29 00:00:00.000","Days In Year":"364"},{"yr":"2010","Last Saturday":"2010-08-28 00:00:00.000","Days In Year":"364"},{"yr":"2011","Last Saturday":"2011-08-27 00:00:00.000","Days In Year":"364"},{"yr":"2012","Last Saturday":"2012-08-25 00:00:00.000","Days In Year":"364"},{"yr":"2013","Last Saturday":"2013-08-31 00:00:00.000","Days In Year":"371"},{"yr":"2014","Last Saturday":"2014-08-30 00:00:00.000","Days In Year":"364"},{"yr":"2015","Last Saturday":"2015-08-29 00:00:00.000","Days In Year":"364"},{"yr":"2016","Last Saturday":"2016-08-27 00:00:00.000","Days In Year":"364"},{"yr":"2017","Last Saturday":"2017-08-26 00:00:00.000","Days In Year":"364"},{"yr":"2018","Last Saturday":"2018-08-25 00:00:00.000","Days In Year":"364"},{"yr":"2019","Last Saturday":"2019-08-31 00:00:00.000","Days In Year":"371"}]}
In this example the year end date is the Saturday that falls closest to the last day of the fiscal year end month.
DECLARE @YEM as int = 8;
SELECT yr,
CAST(YE as datetime) as [YE Date],
YE - PYE as [Days in Year]
FROM
(
SELECT yr,
wct.DATEFLOAT(
yr,
@YEM,
DATEPART(d, wct.DATEFLOAT(yr, @YEM + 1, 1) - 4)
+ wct.Modulo((7 - DATEPART(DW, wct.DATEFLOAT(yr, @YEM
+ 1, 1) - 4)), 7)
) as YE,
wct.DATEFLOAT(
yr - 1,
@YEM,
DATEPART(d, wct.DATEFLOAT(yr - 1, @YEM + 1, 1) - 4)
+ wct.Modulo((7 - DATEPART(DW, wct.DATEFLOAT(yr - 1,
@YEM + 1, 1) - 4)), 7)
) as PYE
FROM
(
VALUES
(2005),
(2006),
(2007),
(2008),
(2009),
(2010),
(2011),
(2012),
(2013),
(2014),
(2015),
(2016),
(2017),
(2018),
(2019)
) n (yr)
) m;
This produces the following result.
yr YE Date Days in Year
----------- ----------------------- ----------------------
2005 2005-09-03 00:00:00.000 371
2006 2006-09-02 00:00:00.000 364
2007 2007-09-01 00:00:00.000 364
2008 2008-08-30 00:00:00.000 364
2009 2009-08-29 00:00:00.000 364
2010 2010-08-28 00:00:00.000 364
2011 2011-09-03 00:00:00.000 371
2012 2012-09-01 00:00:00.000 364
2013 2013-08-31 00:00:00.000 364
2014 2014-08-30 00:00:00.000 364
2015 2015-08-29 00:00:00.000 364
2016 2016-09-03 00:00:00.000 371
2017 2017-09-02 00:00:00.000 364
2018 2018-09-01 00:00:00.000 364
2019 2019-08-31 00:00:00.000 364