Chris Burke wrote:
> I'm happy to include these calculations in dates, and overlooked them
> before.
> 
> However, the ISO 8601 calendar treats the first day of the week as a
> Monday, and I suggest that both wks and weeknumber do likewise. This
> would mean that the left argument defaults to 1. Also, I don't see any
> point in having a different left argument, so I would just make both
> return the ISO 8601 calendar values only. Are there any other ISO
> calculations that return different values?

Just to put a plug in for the astronomical community, who obviously are 
stakeholders when it comes to things related to time:

Please be sure to consult "Astronomical Algorithms" by Jean Meeus 
(Willmann-Bell, 1998, 2nd ed.; 1991, 1st ed.).  Chapter 7 (at least in 
the 1st ed.) deals with the fundamental day counter in astronomy, 
namely, the Julian Day number--and this (and its reverse) is a 
calculation that J should really have built into its dates.ijs script.  
(There's also a Modified Julian Day number, which uses a more recent 
date as its zero point--just as J uses 1/1/1800.)  By the way, the 
Julian Day number is counted from 4713 B.C. (synchronized starting 
point of several astronomical cycles) and is a completely different 
thing than the similar-sounding Julian Date, which is the day number 
within a year starting with January 1.  If you're doing any kind of 
synchronicity of time series data (e.g., agricultural, financial, etc.) 
with astronomical phenomena, the Julian Day number (and Modified Julian 
Day) conversions are essential calculations.

Two other books dealing with time and calendars that might be helpful 
are the following:

"Explanatory Supplement to the Astronomical Almanac" (University 
Science Books, 1992) is one of the foundational books for astronomical 
computing and celestial phenomena--this is an area where J ought to 
shine (pardon the pun)!  Chapter 12 deals specifically with calendars 
and gives some alternative algorithms to the above book for the Julian 
Day number conversions.

"Calendrical Calculations" by Reingold and Dershowitz (Cambridge Univ 
Press, 3rd ed., 2007; 2nd ed., 2001; 1st ed., 1997) is probably the 
premier book available dealing with calendars of all kinds (even a very 
short chapter on the ISO calendar), giving formulas for their 
construction and calculation.

I hope these resources might lead you to include at least some other 
basic date-related verbs.  Thanks for considering them!

Harvey

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to