Re: [Rd] Accuracy of qt for df=1 (PR#9804)

From: <maechler_at_stat.math.ethz.ch>
Date: Wed, 25 Jul 2007 09:57:26 +0200 (CEST)


>>>>> "MW" == Morten Welinder <mwelinder_at_gmail.com>
>>>>> on Tue, 24 Jul 2007 17:18:09 +0200 (CEST) writes:

    MW> Full_Name: Morten Welinder
    MW> Version: 2.2.0
    MW> OS: Linux
    MW> Submission from: (NULL) (216.223.241.229)


    MW> The qt function for df=1 is implemented as...

    MW> q = - tan((P+1) * M_PI_2);

    MW> Adding 1 kills accuracy for P near 0. A better way is to use

    MW> q = cot(P * M_PI_2);
    MW> or
    MW> q = 1/tan(P * M_PI_2);


Yes, indeed, thanks a lot, Morten!!

     ((BTW: you should really upgrade your version of R))

Best regards,
Martin



R-devel_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel Received on Wed 25 Jul 2007 - 08:00:18 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Wed 25 Jul 2007 - 15:36:54 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.