buerstenmacher / hyperprecision

dynamic size arithmetic types

this repository declares and defines two unlimited dynamic integer types and one unlimited floatingpoint type with user defined precision

1.  rom::uintxx_t   an unsigned integer with dynamic and possibly unlimited size
   

2.  rom::intxx_t    an signed integer with dynamic and possibly unlimited size
   

3.  rom::floatxx_t<size_t nr>  (nr defines the number of significant bits of this type)
   

--> download all files into one dirctory

--> compile: g++ example.cpp -o example -std=c++11 -Wno-psabi -Weffc++ -Wall -pedantic -pthread -Os

--> run: ./example //you will see a short demo

example output might be:

nano@nano-desktop:/media/raid/cppwd/hyperprecision$ sudo g++ example.cpp -o example -std=c++11 -Wno-psabi -Weffc++ -Wall -pedantic -pthread -O3 && ./example

e^(0.001) is: +1.0010005001667083e+0

e^(0.01) is: +1.0100501670841680e+0

e^(0.1) is: +1.1051709180756476e+0

e^(1) is: +2.7182818284590452e+0

e^(10) is: +2.2026465794806716e+4

e^(100) is: +2.6881171418161354e+43

e^(1000) is: +1.9700711140170469e+434

e^(10000) is: +8.8068182256629209e+4342

e^(100000) is: +2.8066633604261221e+43429

e^(1000000) is: +3.0332153968020654e+434294

e^(10000000) is: +6.5922325346179653e+4342944

e^(100000000) is: +1.5499767466477715e+43429448

(2^3217- 1) is a mersenne prime number

it is: +259117086013202627776246767922441530941818887553125427303974923161874019266586362086201209516800483406550695241733194177441689509238807017410377709597512042313066624082916353517952311186154862265604547691127595848775610568757931191017711408826252153849035830401185072116424747461823031471398340229288074545677907941037288235820705892351068433882986888616658650280927692080339605869308790500409503709875902119018371991620994002568935113136548829739112656797303241986517250116412703509705427773477972349821676443446668383119322540099648994051790241624056519054483690809616061625743042361721863339415852426431208737266591962061753535748892894599629195183082621860853400937932839420261866586142503251450773096274235376822938649407127700846077124211823080804139298087057504713825264571448379371125032081826126566649084251699453951887789613650248405739378594599444335231188280123660406262468609212150349937584782292237144339628858485938215738821232393687046160677362909315071

e is: +2.71828182845904523536028747135266249775724709369995957496696762772407663035354759457138217852516642742746639193200305992181741359e+0

pi is: +3.14159265358979323846264338327950288419716939937510582097494459230781640628620899862803482534211706798214808651328230664709384460e+0

ln(e^(1.0)) is: +9.9999999999999999e-1

ln(e^(10.0)) is: +.9999999999999999e+1

ln(e^(100.0)) is: +.9999999999999999e+2

ln(e^(1000.0)) is: +.9999999999999999e+3

ln(e^(10000.0)) is: +.9999999999999999e+4

ln(e^(100000.0)) is: +.9999999999999999e+5

ln(e^(1000000.0)) is: +.9999999999999999e+6

ln(ln(ln(e^(e^(e^(7.0)))))) is: +6.9999999999999999e+0 Try this calculation on any other calculator!!!

arcsin(sin(0.8))is: +7.99999999999999999999999999999999e-1

arccos(cos(0.8))is: +7.99999999999999999999999999999999e-1

cos(20)*cos(20) + sin(20)*sin(20) is:

+1.00e-0 [8bit Float]

+1.0000e+0 [16bit Float]

+1.00e-0 [8bit Float]

+1.0000e+0 [16bit Float]

+10.00000000e-1 [32bit Float]

+1.0000000000000000e+0 [64bit Float]

Googol is 10^100 = +9.99999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999e+99

Googolplex is 10^Googol = +9.99999999999999999999999999999999999999999999999999507381353612355094557172415952656787848926056280794561504251052007996099194936e+9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999

Square root of Googolplex is: = +9.99999999999999999999999999999999999999999999999999753690676806177547278586207976328393924463028140397250417984179904556418802957e+4999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999

hyperprecision's Issues

This repository suffers from a serious performance problem. What do you think?