I wrote a leap year function when I was a summer intern in APL (a loooong
time ago). It amazed my more experienced colleagues because it used the
inner product ≠.=
leap←{0≠.=4 100 400∘.|⍵}
leap (100×16+⍳6) ∘.+ ¯5+⍳11
0 1 0 0 0 1 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 1 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
The above translates directly into J:
leap=: 0 ~:/ .= 4 100 400 |/ ]
leap (100*16+i.6) +/ i:5
0 1 0 0 0 1 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
0 1 0 0 0 1 0 0 0 1 0
0 1 0 0 0 0 0 0 0 1 0
On Sat, Sep 27, 2014 at 9:43 PM, robert therriault <[email protected]>
wrote:
> Hi Leslie,
>
> I did say it was a heavy handed method to establish a leap year, so I
> think we are in agreement. :-)
>
> The method that I ended up using was Raul's isLeap which is the J version
> of your algorithm with a clever use of 2 | to ensure that divisible by 4,
> not 100, 400 is the result
>
> isleap=. 2 | +/ 0= 4 100 400 | yt NB. yt is the year of the current date.
>
> By the way if you want to see todayno in J, all you need to do is type in
> its name without an argument and it will return the definition
>
> todayno
> 3 : 0
> 0 todayno y
> :
> a=. y
> if. 1=x do. a=. 0 100 100 #: a end.
> a=. ((*/r=. }: $a) , {:$a) $,a
> if. 1<x do.
> 'y m d h mm s'=. <"_1 |: a
> else.
> 'y m d'=. <"_1 |: a
> end.
> y=. 0 100 #: y - m <: 2
> n=. +/ |: <. 36524.25 365.25 *"1 y
> n=. n + <. 0.41 + 0 30.6 #. (12 | m-3),"0 d
> if. 1<x do.
> n=. n + (24 60 60#.h,.mm,.s)%24*3600
> end.
> 0 >. r $ n - 657378
> )
>
> Also, I believe Linda was correct stating that it has day 0 as being
> January 1st, 1800. This makes it more useful as a way of determining dates
> relative to each other rather than a Gregorian-Julian conversion verb.
>
> todayno 1800 1 1
> 0
>
> Cheers, bob
>
> On Sep 27, 2014, at 9:25 PM, 'Leslie S Satenstein' via Beta <
> [email protected]> wrote:
>
> > leapyear = ((year mod 400) =0 ) or ((year mod 4 = 0) and not ((year mod
> 100)= 0).
> >
> > It seems to me to be excessive use of todayno. Not having seen the code
> to todayno, I would
> >
> > presume it uses a gregorian to Julian day number conversion, The
> julian day is defined as
> >
> > (jan 1 year 1) = zero
> >
> > and todayno calculates the days since.
> >
> > I would presume that the correction includes the 10 day jump in year
> 1580 where the Pope corrected the gregorian calendar by declaring the new
> leapyear rule and .... October 21 1580 was deemed to be Nov 1st of that
> year.
> >
> > Furthermore, if you want to know what day of the week the the 31th of
> the month is, you would realize that it would be 3 days after February 28,
> for non leap years and three days after Feb 29th on leapyears. A similar
> rule applies for months with 30 days.
> > This was an established convention.
> >
> >
> >
> >
> > Regards
> >
> >
> > Leslie
> >
> >
> >
> >
> >
> >> ________________________________
> >> From: robert therriault <[email protected]>
> >> To: Beta forum <[email protected]>
> >> Sent: Wednesday, September 24, 2014 10:28 AM
> >> Subject: Re: [Jbeta] Concern about todayno
> >>
> >>
> >> Hi Henry,
> >>
> >> ...and because it doesn't exist Feb 29, 2003 has the same todayno value
> as March 1, 2003. That seems pretty reasonable to me since checking if
> 2003 2 29 =&todayno 2003 3 1 becomes a way (albeit heavy handed) to
> discern leap years.
> >>
> >> Cheers, bob
> >>
> >> On Sep 24, 2014, at 2:14 AM, Henry Rich <[email protected]> wrote:
> >>
> >>> 2003 2 29 doesn't exist
> >>>
> >>> Henry Rich
> >>>
> >>> On 9/24/2014 5:13 AM, Linda Alvord wrote:
> >>>> Shouldn't the difference be 366 every four years. Bob's comments
> caused me
> >>>> to worry about leap year, and I am getting incorrect answers but they
> may be
> >>>> based on this.
> >>>>
> >>>>
> >>>>
> >>>> diff=: 13 :'(todayno x)-todayno y'
> >>>>
> >>>>
> >>>>
> >>>> (2005 3 1) diff 2004 2 29
> >>>>
> >>>> 366
> >>>>
> >>>> (2004 3 1) diff 2003 2 29
> >>>>
> >>>> 366
> >>>>
> >>>> (2003 3 1) diff 2002 2 29
> >>>>
> >>>> 365
> >>>>
> >>>> (2002 3 1) diff 2001 2 29
> >>>>
> >>>> 365
> >>>>
> >>>> (2001 3 1) diff 2000 2 29
> >>>>
> >>>> 366
> >>>>
> >>>> (2000 3 1) diff 1999 2 29
> >>>>
> >>>> 366
> >>>>
> >>>> (1999 3 1) diff 1998 2 29
> >>>>
> >>>> 365
> >>>>
> >>>> (1998 3 1) diff 1997 2 29
> >>>>
> >>>> 365
> >>>>
> >>>> (1997 3 1) diff 1996 2 29
> >>>>
> >>>> 366
> >>>>
> >>>> (1996 3 1) diff 1995 2 29
> >>>>
> >>>> 366
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> ----------------------------------------------------------------------
> >>>> For information about J forums see
> http://www.jsoftware.com/forums.htm
> >>
> >>
> >>
> >>
> >>>>
> >>> ----------------------------------------------------------------------
> >>> For information about J forums see http://www.jsoftware.com/forums.htm
> >>
> >> ----------------------------------------------------------------------
> >> For information about J forums see http://www.jsoftware.com/forums.htm
> >>
> >>
> >>
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
