EuNumber can do sequence calculations, bigmath, and multiple precision.
Originally written in Euphoria
Uses Euphoria from: https://github.com/jmsck55/euphoria
Get latested updates from: https://github.com/jmsck55/eunumber
You can use this software, but it is experimental. GMP and MPFR are better for digital production software. For analog calculations, such as those that deal with physical matter and light waves, use analog calculating devices. EuNumber is for fast, accurate calculations of digital Big Numbers.
Original Eunumber for fast, accurate calculations of Big Numbers
Copyrighted version.
Experimental.
I don't care about money. I care about saving people's lives.
Details:
GMP and MPFR are better for digital production software. https://www.mpfr.org/
For analog calculations, such as those that deal with physical matter and light waves, use an analog calculating device.
EuNumber is for fast, accurate calculations of digital Big Numbers.
It has some features that can't be done in other big number math libraries.
Use number arays: {1,2,3}
In NewEun(): aEun = NewEun({1,2,3}, 2)
Instead of string literals like: "123"
String literals will not work.
Use instead: aEun = ToEun("1.23e2")
Change your units so that they are close to one (1). If you have 1000 grams, use 1 kilogram for instance. This would be represented as the sequence: {{1} , 0 , maxlength , radix} of type "eun";
Example: 1000 grams would be: {{1} , 3 , maxlength , 10}; or {{1} , 0 , maxlength/3 , 1000};
Zero is represented as: {{} , 0 , maxlength , radix}; notice the empty sequence as the first argument.
2.17 is: {{2 , 1 , 7} , 0 , maxlength , 10};
Note: Maxlength is the maximum number of significant digits. A good maxlength is about 200 or less on modern computers. Always carry out your numbers to a few more decimal places of accuracy, to make sure you have enough accuracy for your number of significant digits.
In other words: Always increase maxlength to a few more integers than you need. This is good programming practise when doing mathematics on a computer.
When doing division, you must observe the "moreAccuracy" variable. All higher math functions rely on division (or "mult_inv(x) = 1/x").
With larger maxlengths, you will need to increase the more accuracy variable. Increase the "more_accuracy" variable to get more accuracy. Decrease it to get less accuracy, with smaller maxlengths.
"moreAccuracy" variable should be about one seventh (1/7) to one sixth (1/6) of maxlength, (1/6 being larger than 1/7).
You can use radix of 2 to 1000 on 32-bit, up to 100,000,000 safely on 64-bit. (It can actually be a little larger than this). I favor preserving accuracy over speed or size, in programming.
Keep in mind: Square roots (even roots) are always plus (+) and minus (-), So you have to account for both plus and minus values, by splitting them up and running both of them through each function or operation that they need to go through. (Complex numbers' imaginary part is a square root, and it, the square root, is always plus and minus.)
Originally written in Euphoria
Uses Euphoria from: https://github.com/jmsck55/euphoria
Get latested updates from: https://github.com/jmsck55/eunumber