[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Prayer time



Nadim Shaikli writes:
 > --- Ahmad Twaijry <msx at hush dot com> wrote:
 > > (I hope if anyone can test them with me and then vote for the best one)
 > 
 > Do consider including other cities outside .sa for those of us not there
 > for further testing (btw: nicely done !!).

Salam, 

I thought we could use some algorithm comparison. This is not an
accuracy test, but more as a way to catch defunct tools and missing
features.


REFERENCE:

My reference is a book called "Al-Mawaqeet wa Al-Qibla" by a Kuwaiti
Astro. pioneer Dr. Saleh Al-Ajairy. The book does include information,
equations, and algorithms for Prayer times, but due to a bad publisher
or copy, must of these equations are useless because the "+ - * /"
signs are not there! However, on page 207, there are some final prayer
times results for Abu-Dhabi on 10-29-1982 (not leap) and the below
comparison will be based on that.


PACKAGES USED:

hdate: the newest version from Debian (testing) 
prayert.c: this is mine and the one that I'm sending to Ahmad now.  
prayer.cpp: this is the one uploaded to the CVS.  
islamicfinder: islamicfinder.org is not a package ;) 


SETTINGS: 

This is the the most important factor in calculating prayer
times. Having the right settings for a certain location is a sure
thing to produce close enough results on any tool. The problem here is
that not all settings and options are supported by these tools. For
example:

** Having different Isha and Fajr angles is not supported by
   "hdate". This is really important.

** Multiple options for Extreme latitudes  is only available in
   "prayert.c"

** Height above sea level is also only available in "prayert.c"

** Ja'fari Fiqh for Al-Assr (and other salats?) is not supported by
   all the tools I've tested.

** Support for Daylight Saving Times is missing from "prayert.c" and
   "prayer.cpp"

Fortunately, the settings used in Al-Ajairy book for Abu-Dhabi are
basic, so we could concentrate on the calculation equations instead. 


To override the default with "hdate" I've used: 
$export PTLOC="AbuDhabi 24 29 N 54 21 E 4 N" 
For other tools lat=24.4833 long=54.35 angles=18 GMT=4


THE RESULTS:

For clarity, a difference of 3-4 minutes from the Al-Ajairy reference
is highlighted with '*'. Differences of 5 minutes or more are
highlighted with a double '**'.

--------------------------------------------------------------------
              Fajr   Shorwwq   Zuhr    `Asr       Maghrib   `Ishaa'
Date Day      Dawn   Sunrise   Noon     Afternoon Sunset    Evening
--------------------------------------------------------------------
islamicfinder 5:11    6:26     12:07    15:20      17:46     19:03*
hdate         5:12    6:30     12:06    15:19      17:42*    19:00
prayert.c     5:11    6:28     12:06    15:19      17:44     19:01
prayer.cpp    5:08    6:24     12:07    15:22      17:49*    19:06**
--------------------------------------------------------------------
Al-Ajairy     5:10    6:26     12:06    15:20      17:46     19:00


AS you can see from the results, there are slight differences but
 nothing major.  almost all results are quite close and all, at least
 in theory, are accurate. The only exception is "prayer.cpp" Ishaa
 results, but that's no biggie.



Regards,

- Thamer Mahmoud