Friday, 12 August 2016

Endless Numeration

endless iteror #2


Frances K*, 2016

Learn to create ever bigger numbers
in an article series on site and blog,
dedicated to Chelsea Manning hero.

© Kreative commons

#2 Numeration

We will unravel how decimal notation works, see how large numbers are handled in history, and explore the boundaries of mathematics.

2.1 Radix notation

The numbers you use daily are in the decimal system. This is a type of radix notation with number base 10 called ten. Although in any radix its base is written as one zero 10.
Here we show numbers in decimal base on an orange background.

You've forgotten (internalized) how this actually works, it's complicated.
The base determines a set of digits in the range 0d<10 to multiply the powers of the base with and add that in series.
This way a word composed of digits equals a unique number n.

Show the popular radices, both in decimal notation (orange) and by simple counting of ones (yellow).


 2 =: 11  binary
 8 =: 11111111  octal
10 =: 1111111111  decimal base
16 =: 1111111111111111  hexadecimal

We use the sign =: to change between formats.
Click @ to put the digits of each radix on the board.

After you agree on a base, apply repeated multiplication to it. Index each position p by a radix power, starting right with 10^0 = 1 and 10^1 = 10 and put larger powers 10^p on the left (Arabic writing direction).

On each numeral place you can write a digit, that multiplies its radix' power. Then add up that series to get your number, as below.


2016 = 1*6+
    10^3*2    in base 10
2015 = 1*7+
     8^3*3  =: 3737  in base 8
2020 = 1*0+
     2^8*1+    in base 2
    2^10*1  =: 11111100100
2018 = 1*2+
          =: 7e2  in base 16

Click @ to represent year numbers in different bases.
Or input your own...
in and


Radix notation is optimal, because we can uniquely express all natural numbers up to any n within a minimum word space. Any other system with a set of characters of radix size will do worse.

This goes for all natural number bases, but real bases are possible too. Try them in the above App: number 10 in base pi for example.
Radices below 2 show increasing overlap as they approach the lower limit of 1, A growing proportion of their digit series expresses the same numbers. So radices in the range (1,2) could model the overlap produced by arithmetical systems with a basic number.

2.2 Trends in history

The mumbo-jumbo of radix notation soon became automated, else society would now be crippled. What luck that children in the ages before social media were able to learn their elementary operations in decimals. We stand on the shoulders of midgets, as well as giants!

The ancient Egyptians had eight numeral hieroglyphs for the powers of ten up to 10000000 that were directly added in any position to depict numbers and baffle the crowd.
The Romans some MDCLXVI years ago had just those seven letters to work with. Yet the digit concept was already present in speech. D our 500 was quingenti in latin, and 8000 horses octo milia equorum.

The left position of decimal digits is most important. But powers of ten elude the human psyche after a few 1000000000 billion.
Bankers who believe Reaganomics will last, choose to ignore the impact of compound interest and the large debt that produces.
Any growth must stop somewhere, but exponential growth rapidly becomes unsustainable.

For example. If a family of two keeps growing at a general annual rate of 1.1% for 2016 years, it will be larger than the current world population. Enter that on a calculator.
2*1.011^2016 7.6E9
Exponent E9 in scientific notation means that you multiply the decimal factor on its left by *10^9 to get the number, or often an approximation.

A percentage sign % is equivalent to an exponent E-2 on the factor. Multiply the current quantity with that number to get the increase. Add the increase to the quantity, to get the next total.

2.3 Physical limits

Radix notation seems more economical than unary notation, but the extra digit signs employed for number input/output do not help the operations in the throughput. If your goal is to generate extreme numbers, without regard for the unnumbered gaps in between: then using just the character 1 is most frugal, an easy win.

Take the tetration 10^^r1 for example. A unary number 1.. is written in exactly that many places. In radix notation this space is a power of 10 smaller, that is 10^^r digit places.

This reduction in word size is sufficient to express only the simplest tetrations, given the resources of our physical universe.


3^^3 = 3^3^3 = 3^27
    =: 111**111111111111111111111111111
     = 7625597484987
4^^3 = 4^4^4 = 4^256
     = 10^(log(4)*256)
     1E154    exponential notation
2^^5 = 2^2^2^2^2 = 2^65536
     = 10^(log(2)*65536)
     E2E4   double exponential notation
3^^4 = 3^7625597484987
     = 10^(log(3)*8E12)
     EEE1    multi-exponential notation

Basically, a power of 10 less subtracts 1 from the tetration iterator.
But even in exponential notation, to approximate moderate input size tetrations, your output becomes too long. Worse is, by reducing your system's precision most numbers will escape the net.

Because we are physical beings we cannot uniquely express the majority of the numbers in the arithmetical sea between the scattered islands of tetration. Even with the help of computers we can only point out a small portion of these illusive numbers in the rather random choice of systems we can make.

According to Seth Lloyd the universe as a computer has at most 10^90 qubits, less than 2^300 binary places. Then the radix representation of tetrations like 4^^4 and 2^^6 by far defeats any known physical data capacity.

The human mind can envision even larger constructs: check out the ancient record 10^^(10^(5*2^120)) in buddhist poetry.
Moving higher, we can still contemplate the strength of our rules, but the big numbers they produce lie wholly beyond imagination…

Berlin 1910, lady mason on a ladder high above a smokey city

No comments:

Post a Comment