Re: mathlib

[Tim Channon <tc@...>]

Here is a real and important usage

"Sinh-arcsinh distributions"
http://biomet.oxfordjournals.org/content/96/4/761.abstract
[Note 1]

Interestingly these tend to ignored even though RA Fisher used
hyperbolics long ago for much the same purpose.

This Lua post might be food for thought
http://lua-users.org/lists/lua-l/2013-01/msg00432.html

Lua is hardly orthogonal, good, that is a pedant's game but also squeaky
wheels get out so development pressure tends to push projects in the
direction of a majority want.

A thought.
If Lua contained a numeric array type where the internal representation
was native binary array it is a trapdoor to acting as a controller for
numeric processing. Perhaps in a way able to act more like dedicated
tools, which always have other compromises.

Note 1: one of the schemes is trivially simple, here as a spreadsheet
can do this on bipolar data, facilitating normalisation of skew and
kurtosis, the latter is very rarely done.

SINH( kurtosis_constant * ASINH(in_value) - skew_constant)
  where neutral effect is
  kurtosis_constant = 1 and skew_constant = 0

On 03/04/2014 21:35, Roberto Ierusalimschy wrote:
> We are considering removing some functions from the standard math lib,
> either because we think few people use them or because they are trivially
> implemented without the library. The current list is this:
> 
> - sinh, cosh, tanh: (They are quite specialized, on par with several
> other functions offered by external libraries, such as lhf's mathlibx.)
> 
> - deg, rad, pow: trivially done without the library.
> 
> Comments?
> 
> -- Roberto
> 
>