From madrigal at kundor.org  Fri Dec 11 00:30:07 2009
From: madrigal at kundor.org (Nick Matteo)
Date: Fri, 11 Dec 2009 00:30:07 -0500
Subject: [OpenMadrigal-users] strangeness in getFyear
Message-ID: <eecf74b40912102130i3116ff3chac044bd4afd261e3@mail.gmail.com>

Hello,
I was comparing various time parameters available from Madrigal.  So I
computed the fractional year from the 'year', 'dayno', 'hour', 'min',
and 'sec' parameters:
fyear = year + (dayno - 1 + (hour + (min + sec/60)/60)/24)/366
(this was for 2008, hence 366.)
I kept seeing a .0015 discrepancy, which is about 13 hours.  I thought
I'd check the source to see if fractional year is defined in a
different way.
The relevant computation seems to be in
source/madc/madrec/madDeriveMethods.c, getFyear.
I found these lines:

fyear = (double)iyr;
daysNow = ((double)(dayOfYear - 1));
daysNow += ((double)(24 * hour + min))/(24.0*60.0);
daysNow += ((double)ics)/ (100.0*3600.0*24.0);
fyear +=  daysNow / ((double)daysInYear);

(24*hour + min)/(24*60), or in other words hour/60 + min/(24*60),
doesn't seem right.  Surely it should be hour/24 + min/(24*60)?  Or,
equivalently, (hour + min/60)/24?

Thanks,
Nick

From brideout at haystack.mit.edu  Wed Dec 23 09:19:48 2009
From: brideout at haystack.mit.edu (Bill Rideout)
Date: Wed, 23 Dec 2009 09:19:48 -0500
Subject: [OpenMadrigal-users] strangeness in getFyear
In-Reply-To: <eecf74b40912102130i3116ff3chac044bd4afd261e3@mail.gmail.com>
References: <eecf74b40912102130i3116ff3chac044bd4afd261e3@mail.gmail.com>
Message-ID: <1721A43B-A535-4381-AA39-B2F797EEAD49@haystack.mit.edu>

Nick,

You are correct.  I've patched our Madrigal server here at Millstone  
Hill, but the bug will still appear at the other Madrigal sites until  
the next release.  Thanks for pointing out the problem, and tracing it  
all the way into the C code!

Bill Rideout

On Dec 11, 2009, at 12:30 AM, Nick Matteo wrote:

> Hello,
> I was comparing various time parameters available from Madrigal.  So I
> computed the fractional year from the 'year', 'dayno', 'hour', 'min',
> and 'sec' parameters:
> fyear = year + (dayno - 1 + (hour + (min + sec/60)/60)/24)/366
> (this was for 2008, hence 366.)
> I kept seeing a .0015 discrepancy, which is about 13 hours.  I thought
> I'd check the source to see if fractional year is defined in a
> different way.
> The relevant computation seems to be in
> source/madc/madrec/madDeriveMethods.c, getFyear.
> I found these lines:
>
> fyear = (double)iyr;
> daysNow = ((double)(dayOfYear - 1));
> daysNow += ((double)(24 * hour + min))/(24.0*60.0);
> daysNow += ((double)ics)/ (100.0*3600.0*24.0);
> fyear +=  daysNow / ((double)daysInYear);
>
> (24*hour + min)/(24*60), or in other words hour/60 + min/(24*60),
> doesn't seem right.  Surely it should be hour/24 + min/(24*60)?  Or,
> equivalently, (hour + min/60)/24?
>
> Thanks,
> Nick
> _______________________________________________
> OpenMadrigal-users mailing list
> OpenMadrigal-users at openmadrigal.org
> http://www.openmadrigal.org/mailman/listinfo/openmadrigal-users
>
> !DSPAM:4b321fd032161687619800!
>