comp.lang.idl-pvwave archive: archive » IDL is not accurate enough!

Home » Public Forums » archive » IDL is not accurate enough!

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: IDL is not accurate enough! [message #62558 is a reply to message #62192]

Mon, 15 September 2008 06:40

pgrigis
Messages: 436
Registered: September 2007

Senior Member

noahh.schwa...@gmail.com wrote:
> On 11 sep, 16:44, pgri...@gmail.com wrote:
>> pgri...@gmail.com wrote:
>>> noahh.schwa...@gmail.com wrote:
>>>> On 28 ao t, 18:42, F LDY Lajos <fo...@rmki.kfki.hu> wrote:
>>>> > On Thu, 28 Aug 2008, noahh.schwa...@gmail.com wrote:
>>>> > > Hi,
>>
>>>> > > I've been having problems with IDL accuracy. I'm trying to perform
>>>> > > calculations using the gamma function. The problem is that it grows
>>>> > > VERY fast! Performing this calculation in double (namely gamma(x)/
>>>> > > gamma(y) with x and y big) yields the result: NaN...
>>>> > > Would it be possible to use a program like 'Mathematica' (or any
>>>> > > other) and to plug it in my ILD program? Some kind of CALL_EXTERNAL
>>>> > > that is to say. If it is possible, how can I do it and what is the
>>>> > > best program to use?
>>
>>>> > > Thanks,
>>>> > >Noah
>>
>>>> > gamma(x)/gamma(y) => exp(lngamma(x)-lngamma(y))
>>
>>>> > regards,
>>>> > lajos
>>
>>>> lngamma works fine for my propose! Would you know if an equivalent
>>>> function exists for the beselk function? Something like lnbeselk?
>>>> beselk(x) for x>709 doesn't seen to work.
>>
>>> Isn't 0 a good enough approximation?
>>
>> If not, log(K(x,n))~ln(sqrt(!pi/(2*x)))-x for large x
>>
>> Paolo
>>
>>
>>
>>> Paolo
>>
>>>> If not, I guess that I'll have to wait for the DLMs that add arbitrary
>>>> precision floating point...
>>
>>>> cheers,
>>>> Noah
>
>
>
> Hi Paolo,
> Your approximation seems to be missing a factor? This is what IDL
> gives me:
>
> IDL> x=705d & n=1.1 & print, alog10(beselk(x,n)), (alog(sqrt(!pi/
> (2*x)))-x)
> -307.50372 -708.05331

I meant the natural log (why should a bessel function
care about base 10 anyway?), so use alog instead.

Cheers,
Paolo

>
> Cheers,
> Noah

Report message to a moderator

[Message index]

		IDL is not accurate enough! By: noahh.schwartz on Thu, 28 August 2008 09:30
		Re: IDL is not accurate enough! By: Haje Korth on Tue, 02 September 2008 05:20
		Re: IDL is not accurate enough! By: Kenneth P. Bowman on Fri, 05 September 2008 09:53
		Re: IDL is not accurate enough! By: MP on Fri, 05 September 2008 09:18
		Re: IDL is not accurate enough! By: Michael Galloy on Fri, 05 September 2008 08:57
		Re: IDL is not accurate enough! By: Kenneth P. Bowman on Thu, 04 September 2008 13:33
		Re: IDL is not accurate enough! By: pgrigis on Thu, 11 September 2008 07:44
		Re: IDL is not accurate enough! By: noahh.schwartz on Mon, 15 September 2008 02:37
		Re: IDL is not accurate enough! By: pgrigis on Thu, 11 September 2008 07:35
		Re: IDL is not accurate enough! By: pgrigis on Mon, 15 September 2008 06:40

Previous Topic:	Re: Maximum index for arrays?
Next Topic:	Re: IDL calls C calls IDL?

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Oct 12 12:22:18 PDT 2025

Total time taken to generate the page: 1.91869 seconds