[cfp-interest 3594] Re: csinh

[Jim Thomas]

> On Sep 11, 2025, at 7:59 PM, Fred J. Tydeman <tydeman at tybor.com> wrote:
> 
> C23: G.6.3.5 The csinh functions has these two requirements:
> 
> - csinh(+0 + iNaN) returns ±0 + iNaN (where the sign of the real part of the result is unspeci-
> fied).
> 
> - csinh(NaN + i0) returns NaN + i0.
> 
> Since csinh( x + I*y ) == sinh(x)*cos(y) + I*cosh(x)*sin(y)

> 
> It seems to me that the last case should be:
> 
> - csinh(NaN + i0) returns NaN +/- i0
> (where the sign of the imaginary part of the result is unspeci-
> fied).

For the csinh(NaN + i 0) case, cosh(x) is positive for any numerical value of x and the imaginary part cosh(x)*sin(+0) is +0 for any finite value of x. 

The current specification is consistent with viewing csinh(x + i 0) as a function mapping the edge of the upper half plane to itself. And it avoids an unspecified sign. 

This is unlike the case for csinh(+0 + i NaN), because the real part sinh(+0)*cos(y) varies between +0 and -0 for finite y.

- Jim Thomas

> 
> 
> ---
> Fred J. Tydeman        Tydeman Consulting
> tydeman at tybor.com      Testing, numerics, programming
> +1 (702) 608-6093      Vice-chair of INCITS/C (ANSI "C")
> Sample C17+FPCE tests: http://www.tybor.com
> Savers sleep well, investors eat well, spenders work forever.
> 
> _______________________________________________
> cfp-interest mailing list
> cfp-interest at oakapple.net
> http://mailman.oakapple.net/mailman/listinfo/cfp-interest

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20250912/4768ec24/attachment.htm>