Hyper (precision) numeric data type

[iZ!]

I admit, this one was kinda shitty and it wasn't fully functional.
But now this one is the real thing. Nearly...

Hit me up @ https://github.com/SilverfoxSystems/Hyper
& http://silverfox.systems/Hyper/

[HSE]

What numeric type is that?

[iZ!]

It's a mixture of a String and a floating-point number. It consists of Int64 array and an Int32 which describes the lowest exponent.

EDIT:
"Adaptive float" would maybe describe it best

[HSE]

What are range and precision of that format?

[iZ!]

As far as the mentioned Int64 array's size goes, this is how big the mantissa can be.

I.E.
if initialized as

Code Select Expand

New Hyper(5000,-5000) (low and high "exponents")
then it'd have buffer(mantissa) of length of 10000 * 8 bytes.

The Int32 mentioned describes number of starting QWORD or as I call it, Exponent64 or exp64 in my descriptions.
In fact, the number used is called precision which is the negated LOWEST exp64 .

So the precision is enormous - as many bits there are in approx. 8 * 2GB with the mentioned Int32 offset.
You can have an entire RAM describing a single number :)

[HSE]

Ok. It's an "arbitrary precision" format.

Precision depends on mantissa, not exponent. Range depends on exponent.

 

[iZ!]

i agree    never heard of that format, thanx 4 clearing it out

only thing that maybe differs from others is that it works on balanced numeral system (with base 2^64) - some advantages: sign bit isn't needed, every 64th bit is negative => no rounding error + no need of filling bits when expanding mantissa's size

[jack]

@ i Z !
the one thing that bothers me is the restrictive license

[iZ!]

@ i Z !
the one thing that bothers me is the restrictive license

it lets you do whatever with it. only not making money off my hard work :)

or you had something specific in mind?

[jack]

or you had something specific in mind?

not really, I also have dabbled in writing my own multi precision arithmetic routines, just for fun.
what number base do you use for your routines? 

[jack]

i Z !
this the kind of license that would acceptable to me, it's from https://bellard.org/libbf/

/*
 * Tiny arbitrary precision floating point library
 *
 * Copyright (c) 2017-2020 Fabrice Bellard
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */

[iZ!]

what number base do you use for your routines? 

it works on balanced numeral system (with base 2^64)

[iZ!]

* Permission is hereby granted, free of charge, to any person obtaining a copy...
...
...
...

Sorry but I get lost before finishing reading it 

EDIT this is a MIT licence (according to website), which allows commercial use, so i'm not alright with that

[jack]

suit yourself, MIT is one the most simplest licenses, if you can't understand that then I doubt your ability to code a decent MP library

[iZ!]

suit yourself, MIT is one the most simplest licenses, if you can't understand that then I doubt your ability to code a decent MP library

hah, give it a go mr lawyer

it's as 'decent' as it gets :)