s-news
[Top] [All Lists]

Re: DIGITS

To: MAGGIO GUIDO <GUIDO.MAGGIO@ESTERNI.GRUPPO.MPS.IT>
Subject: Re: DIGITS
From: P Ehlers <ehlers@math.ucalgary.ca>
Date: Mon, 19 Dec 2005 07:07:42 -0700
Cc: s-news@lists.biostat.wustl.edu
In-reply-to: <07B40C0AFE0DBE4FB7D979FF3135A92C012AC69A@se000010010035.servinternet.local>
References: <07B40C0AFE0DBE4FB7D979FF3135A92C012AC69A@se000010010035.servinternet.local>
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax)
Have you considered logarithms?

Peter Ehlers
U of Calgary

MAGGIO GUIDO wrote:

Can anybody tell me how I can make computations of a large number? Example: I am computing a likelihood, within its expression I have
lambda^(n*mean(x)) which in my case is equivalent approximately to
14,4^919. Splus can't make this computation (it could make 14,4^799) as
the result is too high and becomes infinity ("inf"). This causes and impossibility to compute the likelihood.
I have tried enlarging the number of digits through

OPTIONS==>GENERAL SETTINGS==>COMPUTATIONS
And by setting the number of digits to the maximum=20.
This allows me to compute very small numbers with many digits bit does
not help me with my lambda^(n*mean(x))  which again turns out =inf.

I am using Splus 6.2. Is this the cause of my problem?
Thanks

Guido





<Prev in Thread] Current Thread [Next in Thread>
  • DIGITS, MAGGIO GUIDO
    • Re: DIGITS, P Ehlers <=