From: "Bradley Lucier" <xxxxxx@github.com>
Date: Jun 18, 2017 8:03 AM
Subject: Re: [scheme-requests-for-implementation/srfi-144] Added portable sample implementation. (#1)
To: "scheme-requests-for-implementation/srfi-144" <srfi-144@noreply.github.com>
Cc: "Subscribed" <xxxxxx@noreply.github.com>
Will: I believe the following behavior is correct:
> (map (lambda (f) (f 1e-20)) (list flsinh fltanh flasinh flatanh))
(1e-20 1e-20 1e-20 1e-20)
> (map (lambda (f) (f 1e-10)) (list flsinh fltanh flasinh flatanh))
(1e-10 1e-10 1e-10 1e-10)
Your implementation returns
> (map (lambda (f) (f 1e-20)) (list flsinh-144 fltanh-144 flasinh-144 flatanh-144))
(0. 0. 0. 0.)
> (map (lambda (f) (f 1e-10)) (list flsinh-144 fltanh-144 flasinh-144 flatanh-144))
(1.000000082740371e-10
1.000000082740371e-10
1.000000082690371e-10
1.000000082640371e-10)
I'm remarking only because you go to the trouble to provide Taylor series implementations for arguments close to zero for some functions. These functions could use something similar (or to express them in terms of flexp-1 or fllog1+, where you do give a careful Taylor series for small arguments).
In general, it's not good to have values of x for which |asinh(x)|>x or |atanh(x)|>1.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.