Enjoy forums? Start your own community for free.
zIFBoards - Free Forum Hosting
Welcome to Dozensonline. We hope you enjoy your visit.
You're currently viewing our forum as a guest. This means you are limited to certain areas of the board and there are some features you can't use. If you join our community, you'll be able to access member-only sections, and use many member-only features such as customizing your profile, and sending personal messages. Registration is simple, fast, and completely free. (You will be asked to confirm your email address before we sign you on.)
Join our community!
If you're already a member please log in to your account to access all of our features:

Name:   Password:


Pages: (10) « First ... 8 9 [10]  ( Go to first unread post )

 Le Tour Des Bases, Visit each number base; try them out
Double sharp
Posted: Nov 7 2017, 04:24 AM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



I suppose the "ugly" sort of semitotative would involve such concoctions as digit-478 base 720, which is 2 * 239. This is of course the most common sort; transparent (e.g. decimal 6) or semitransparent (e.g. dozenal a) totatives are very rare because the neighbours can't ameliorate everything past a certain point. The first time we get a really bad semitotative is hexadecimal e, which is opaque; the first time we get a not only bad but also fairly useless one is probably tetravigesimal m.

The main issue is that there's a difference between single digits and numbers above "10". The former group are simple elements that have to be manipulated, at least if we for simplicity think only about pure bases. The latter group are meant to be broken down additively. It doesn't really matter that 478 is a semicoprime number in decimal because we can work with it simply as 400 + 70 + 8. But when it is a digit, it does matter.

The need to break down numbers to visualise them past a certain point seems to be so innate that I really doubt a base higher than 20 to 30 could ever fly. Even in that range, I presume there would be a lot of sub-base thinking, and the quadratically increasing size of the multiplication table would push them over the line for actual use. The "break" at 30, and the falling out of 24 when we consider regulars as well as divisors, seems to indicate that mathematical considerations are at one with psychological considerations in telling us to go back down the mountain already at about 18 or 20.
Top
Double sharp
Posted: Nov 7 2017, 04:25 AM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



Why does the "missing last post" bug keep happening? It happened at the start of page 9 and again at the start of page 10. Hopefully this post will fix it.

EDIT: Yes, it did. But can we perhaps look into a more permanent solution that doesn't involve this sort of vacuous post?
Top
wendy.krieger
Posted: Nov 7 2017, 08:12 AM


Dozens Demigod


Group: Members
Posts: 2,432
Member No.: 655
Joined: 11-July 12



I'm looking at a rexx script to make the tables, but the sample table (300) seems to have some errors in it. At the moment, i am writing the story-boards, and it looks like this.

CODE

extproc weave.cmd
goto :eof
!topic about

A program to produce the digit-tables as shown on DozenOnline, one for the digits
one for the intuitive division test.

!src
!inc program;  Here is the proggie.
!end

!topic - Numbers

Relative to a base, numbers have a co-composite and a co-prime part.  This is the
regular bit and the period bit,

  coprime classes
     u    coprime = 1 (ie regular)
     a    coprime divides alpha or A2
     w    coprime divides omega or A1
     y    coprime divides A1A2 = R2
     e    other coprimes
  cocomposite class
     n    cocomposite = 1
     r#   cocomposite divides b^#
     s#   cocomposite divides b^{1#},
  general class
     p    prime number
     c    composite number NSFA
     q    number contains a square factor
     x    number is a proper power

No letter is used twice, so it's possible to grep for inclusion or exclusion of a
particular letter.

!topic - Digits of Base $base

The style is to calculate the class-string for each number to the base, and then print
the two tables separately.

                         Digits           Divisibility
     un     unit,        purple           purple
     ur1    divisor      red              red
     ur2    ring 2       pink             shade of red.
     ur3    ring 3        ,,                 ,,
     wn     div A1
     an     div A2
     yn     div R2
     en     coprime
     wr     cp | A1
     ar     cp | A2
     yr     cp | R2
     er     totitive

For the coprimes, there is some sort of shading by grey, as follows.  We either do this
the hard way, or use f1103 to do the hard work.  This would entail constructing a vector
for the base, and then dotting it against successive rows of f1103.  It also means that
we would have to distribute f1103 to users.

    en, i=1    dark grey
    en, i=2    mid grey
    en, i>2    light grey

Since the intent is to evodently find the long periods vide the other ones that the GQR
finds, we could change i=2 into i | 2x

   select
      when i = 1 then dark-grey
      when 2x // i = 0 then mid-grey
      otherwise light-grey

!topic - Mežod of attack.

  Since we have to get the ring of the regular, and the coprime, the easiest way
  is to do successive GCD until 1 is the gcd

  eg     b = 120  n = 7168  gcd = 8  ring = 0
         b =   8  n =  896  gcd = 8  ring = 1
         b =   8  n =  112  gcd = 8  ring = 2
         b =   8  n =   14  gcd = 2  ring = 3
         b =   2  n =    7  gcd = 1  ring = 4

  r4

  Having found this, we then do

   select
      when 7 equals 1      then type = u
      when 7 divides 120-1 then type = w
      when 7 divides 120+1 then type = a
      when 7 divides 120*120-1 then type = y
      otherwise  type = o

   r4w   1/7168 = 0:00 02 01 08 68 68 68 68 68 has 4 leads and a period of 1.

   The co-composite is 7168/7 = 1024,

This happens without having to do any factorisations.

!topic - General Class

This tells us about the number.

  p  Number is prime
  c  Number is composite NSFA
  q  Number contains a square factor
  x  Number is a proper power.



Top
wendy.krieger
Posted: Nov 7 2017, 08:20 AM


Dozens Demigod


Group: Members
Posts: 2,432
Member No.: 655
Joined: 11-July 12



I'm still not sure about how to proceed with the primes.

Over 3/4 of the primes have a period that is derived from Gauss's rule, split between the long, short in slightly equal proportions. Of the non-gaussian primes, we could simply list these at the foot of the table.

In short, it's a matter of deriving the reduced index, which is varying 1, 2, or something bigger. The period of the prime is (p-1)/i. We can accurately uses gauss's rule to find the value is even or odd, and for proper powers, if i | 2x, (where x is the exponent), then the gaussian-fermat rules suffice. For example, for base 16, x = 4, so we reject any value where i divides 8. The first instance is 31, where i=6, and we seek a division into 30 gives a period of 5, rather than the default double-square value of 15.

Top
Double sharp
Posted: Nov 7 2017, 08:23 AM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



The different shades of grey for the coprimes indicates the length of their periods, IIRC.

EDIT: Yes, darkest grey is period (p-1), medium grey is period (p-1)/2, and lightest grey is anything shorter (except of course periods 1, 2, and 4, which are coloured as omega, alpha, square-omega, or square-alpha).
Top
wendy.krieger
Posted: Nov 7 2017, 08:46 AM


Dozens Demigod


Group: Members
Posts: 2,432
Member No.: 655
Joined: 11-July 12



I was looking at your '300' table, and noticed that neither 91 nor 161 were coloured in some R2 colour. (ie alpha, omega, or the product).

What do you do with composites in general class.

Pouring over icarus's output, i can read most of it. I noticed he had a yates-table in there, that is, primes with periods 1,2,3,4,5... places. I use a different order here, the cunningham style, or true-size. For this process, i will use cunningham, which list the odd numbers just before the double-odd, eg 1,2,4,3,6,8,5,10,12.

At home, i use true-size tables, viz 1,2,6,4,3,10,12,8,5,14,18,9,7.

CODE


          Yates          Carmichael        Size
    1            R      1          R    1         R
    2           11      2         11    2        11
    3          111      4        101    6        R1
    4          101      3        111    4       101
    5        11111      6         R1    3       111
    6           R1      8      10001   10      R0R1
    7      1111111      5      11111   12      RR01
    8        10001     10       R0R1    8     10001
    9      1001001     12       RR01    5     11111
   10         R0R1      7    1111111   14    R0R0R1
   11  11111111111     14     R0R0R1   18    RRR001
   12         RR01     16  100000001    9   1001001



The carmichael system gives an easy to find but well-sorted sizing of the algebraic roots. I've produced an size-sort as far as everything less than 164 digits, but outside of grep, it's rather hard to find the factors. They go quite unimaginably large, but the latest version of factor, seems to handle 120-digit numbers quite well.

Top
Double sharp
Posted: Nov 7 2017, 09:51 AM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



Eh? 91 and 161 are coloured light blue as alpha-omega mixes. Also, it's Icarus' table, not mine.
Top
icarus
Posted: Nov 7 2017, 12:12 PM


Dozens Demigod


Group: Admin
Posts: 1,913
Member No.: 50
Joined: 11-April 06



Note: I have had to change my "limits" post, since 2310 is semitotative-dominant, and that is more telling than being highly semitotative.
Top
Double sharp
Posted: Nov 7 2017, 03:23 PM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



I have filled out base 252. I think for now I will be doing one a day - so Icarus is definitely pulling ahead, which I shall indeed be most grateful for! happy.gif
Top
icarus
Posted: Nov 7 2017, 04:22 PM


Dozens Demigod


Group: Admin
Posts: 1,913
Member No.: 50
Joined: 11-April 06



I am coding the paragraph. It's just easier & faster that way, bite the bullet. Now it won't take into account grammar for prime powers. The problem with coding language is all the which-switch statements for grammar, oxford commas, etc. and you can get really awkward robot talk. Some of the prose is dry, but I figure it's a good starting point and better than my plucking around for all the pieces.

Now I can also get into nitty gritty and count opaque totatives, opaque semitotatives, alpha semitotatives, etc. I figure if you're using the first paragraph as is, I can update it to add the gobbledygook. I also thought of making a little chart of it, but that will require more coding and delay everything.

Here's a test of what I have right now. I am going to be madly interrupted all day.

Base 432 | Quadringentoduotrigesimal (Dozenal: Base 300; | Trinabiqual)

The number 432 is neither squarefree nor a prime power with 2 distinct prime divisors {2, 3}. Its prime decomposition is 24 × 33 and has the prime signature “43”, the smallest base to have such signature. Four hundred thirty two number. Four hundred thirty two has 20 divisors {1, 2, 3, 4, 6, 8, 9, 12, 16, 18, 24, 27, 36, 48, 54, 72, 108, 144, 216, 432}, 20 regular numbers in all. Base 432 has 144 totatives (1 out of 3 digits, 33.3%), and there are 81 prime totatives. The quadringentoduotrigesimal ω-number is 431, a prime.. The α-number is 433, also prime. Both ω and α prime makes for a “bone dry“ set of quadringentoduotrigesimal intuitive divisibility tests that cover 2 of the 8 smallest primes. There are 269 neutral digits, of which 12 are semidivisors (richness 2: 32, 64, 81, 96, 128, 162, 192, 243, 256, 288, 324, 384) and 257 are semitotatives (10, 14, 15, 20, 21, 22, 26, 28, 30, 33, 34, …). Of the 257 semitotatives, there are 28 alpha-2-related semitotatives and 229 opaque semitotatives. Long primes in base 432 include 5, 7, 17, and 19. In base 432, 5 is a third-order Wieferich prime. There are 44 quadringentoduotrigesimal quadratic residues.

Base 210 | Ducentodecimal (Dozenal: Base 156; | Unpenthexal)

The number 210 is squarefree with 4 distinct prime divisors {2, 3, 5, 7}. Its prime decomposition is 2 × 3 × 5 × 7 and has the prime signature “1111”, the smallest base to have such signature. Two hundred ten is a primorial that also is the seventeenth highly regular and the thirteenth highly semidivisible number. Two hundred ten has 16 divisors {1, 2, 3, 5, 6, 7, 10, 14, 15, 21, 30, 35, 42, 70, 105, 210}, 16 regular numbers in all. Base 210 has 48 totatives (8 out of 35 digits, 22.9%), and there are 42 prime totatives. The ducentodecimal ω-number is 209, which is divisible by 11 and 19. The α-number is 211, a prime. The base has omega-dominant intuitive divisibility tests that cover the 5 smallest and further, 6 of the smallest 8 primes. There are 147 neutral digits, of which 52 are semidivisors (quantities by richness starting with 2: 25, 14, 7, 3, 2, 1) and 95 are semitotatives (22, 26, 33, 34, …). Of the 95 semitotatives, there are 23 omega-related semitotatives and 72 opaque semitotatives. Long primes in base 210 include 13 and 17. There are 48 ducentodecimal quadratic residues.

Base 120 | Centovigesimal (Dozenal: Base a0; | Decanunqual)

The number 120 is neither squarefree nor a prime power with 3 distinct prime divisors {2, 3, 5}. Its prime decomposition is 23 × 3 × 5 and has the prime signature “311”, the smallest base to have such signature. One hundred twenty is a that also is the fifth superior highly composite, tenth a highly composite, and the fourteenth highly regular number. One hundred twenty has 16 divisors {1, 2, 3, 4, 5, 6, 8, 10, 12, 15, 20, 24, 30, 40, 60, 120}, 16 regular numbers in all. Base 120 has 32 totatives (4 out of 15 digits, 26.7%), and there are 27 prime totatives. The centovigesimal ω-number is 119, which is divisible by 7 and 17. The α-number is 121, the square of 11. Base 120 thus has a relatively rich set of intuitive divisibility tests that cover the 5 smallest and further, 6 of the smallest 8 primes. There are 73 neutral digits, of which 20 are semidivisors (richness 2: 9, 16, 18, 25, 32, 36, 45, 48, 50, 64, 72, 75, 80, 90, 96, 100; richness 3: 27, 54, 108; richness 4: 81) and 53 are semitotatives (14, 21, 22, 26, 28, …). Of the 53 semitotatives, there are 8 alpha-related semitotatives, 16 omega-related semitotatives, and 29 opaque semitotatives. In base 120, 11 is a second-order Wieferich prime. There are 18 centovigesimal quadratic residues (0, 1, 4, 9, 16, 24, 25, 36, 40, 49, 60, 64, 76, 81, 84, 96, 100, and 105).



Note: I recognize the grammar hiccup, will try to correct so long as it doesn't eat up all the time; moving on to add the other components, but like I said, going to be interrupted as heck today

201711071418: Nearly through but have to drop for the day. Some glitches with a Which statement, but will get through and move ahead this evening or tomorrow. Remaining: list semidivisors, clean up some language, and add the Wieferich, long primes, quadratics. Then things may go more smoothly, since I won't be composing prose each time, but instead have the machine write it directly from the register program. This interpreter is a page long itself; the register is several pages long and the flexCell is longer; all in Wolfram, famous for brief code!

201711072210: Finished it!! Now it should glide on freakin' rails.
Top
Double sharp
Posted: Nov 7 2017, 11:47 PM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



I'm not actually using the first paragraph as is, but I am certainly using the information from it, as you can see in the latest tours, and spreading it among the "vital statistics" bullet points.
Top
icarus
Posted: Nov 8 2017, 04:24 AM


Dozens Demigod


Group: Admin
Posts: 1,913
Member No.: 50
Joined: 11-April 06



The cutting and pasting of HTML from the automation means that I can put up a summary in about 5 minutes. The thing that takes the most time is writing the tags for the "Digit Map", etc.! But it is time to turn in. Will hit many more tomorrow. The algorithm isn't perfect, but I don't want to delay things by "fixing" it.
Top
Double sharp
Posted: Nov 8 2017, 04:49 AM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



QUOTE (icarus @ Nov 8 2017, 04:24 AM)
The cutting and pasting of HTML from the automation means that I can put up a summary in about 5 minutes. The thing that takes the most time is writing the tags for the "Digit Map", etc.! But it is time to turn in. Will hit many more tomorrow. The algorithm isn't perfect, but I don't want to delay things by "fixing" it.

Looking forward to it!

I think the automation works best for bases above 120 or maybe 168 because then you can nearly completely forget about the potential civilisational use at any level of civilisation: there simply is none. For bases up to 120 we need to worry about all those tricks like AMT and mixed radices for getting things done.

I guess AMT production can be automated, but I am not sure if its practicality can be. Is the AMT for base 60 practical? Certainly. Is the AMT for base 117 practical? Certainly not. Is the AMT for base 120 practical? It's getting a bit big and Wendy's alternating arithmetic seems better, but I suppose we could handle this simply by declaring that the table shall not have more products than the full multiplication table of hexadecimal or maybe vigesimal. Is the AMT for base 105 practical? It contains a splendid quantity of complementary divisors to leverage all of the first four primes except the one we actually want, 2. Is the AMT for base 21 practical? It gives hardly any leverage, but it is very compact!

For this reason I am focusing on the grand bases first, as they can be done quite a bit more simply. But I haven't forgotten the mid-scale ones: I will soon do up 50, and then I plan to add a bunch of interesting odd bases: {45, 63, 65, 75, 105, 117}. (Feel free to scold me if you think there are some more interesting ones that I am about to unjustly neglect. I know Wendy has mentioned 165 and 195 as having abundant regulars but I don't see why they'd be much different from 105, to be honest, and apart from the detailed approximation for pi, 113 seems totally useless as a large prime.)
Top
icarus
Posted: Nov 8 2017, 12:40 PM


Dozens Demigod


Group: Admin
Posts: 1,913
Member No.: 50
Joined: 11-April 06



Double sharp:

The AMT is automated already. It needs several not-necessarily contiguous small factors less than the square root of the base. How many is not an exact science, but can easily be codified such that we might write code to throw it when necessary. 60, for example, is ideal in that its divisors are neatly cleaved about its square root (as all bases have) with 6 contiguous smallest factors (not all bases have this). That is why AMT is feasible there. AMT is not really applicable to bases the size we are entertaining at the aggregator thread. I do have a prime map like you see in the old school threads, etc.

Automated example for base three dozen:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
3 6 9 12 15 18 21 24 27 30 33 36            
4 8 12 16 20 24 28 32 36                  
5 10 15 20 25 30 35                      
6 12 18 24 30 36                        


I do have some segments written. Divisors, etc. What I had been working on is getting the tables to work directly from the register. This will make them far more flexible than flexCell. Also fully automated is the entire intuitive divisibility test section. (A number x is divisible by 2 if...) This section even incorporates the notion of "trine, nontrine, overtrine, undertrine" (I never want to be undertrine lzozllz), so accommodates nondecimal bases divisible by three.

Example of autogenerated intuitive tests, using base 126:

An arbitrary integer x is divisible by …

  • 2 if the least significant place value of x is one of {0, 2, 4, 6, 8, …, 124} (63 combinations). (Regular test, richness = 1: divisor).
  • 3 if the least significant place value of x is one of {0, 3, 6, 9, 12, …, 123} (42 combinations). (Regular test, richness = 1: divisor).
  • 4 if the 2 least significant place values of x are one of {0000, 0004, 0008, 0012, 0016, ...., 5872} (3969 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is even and the last place value is one of {00, 04, 08, 12, 16, .., 24},
    or if the penultimate place value is odd and the last place value is one of {02, 06, 10, 14, 18, .., 22}, then x is divisible by 4.
  • 5 if the centohexavigesimal digital root of x is divisible by 5. (Neighbor factor test: omega).
  • 6 if the least significant place value of x is one of {0, 6, 12, 18, 24, …, 120} (21 combinations). (Regular test, richness = 1: divisor).
  • 7 if the least significant place value of x is one of {0, 7, 14, 21, 28, …, 119} (18 combinations). (Regular test, richness = 1: divisor).
  • 8 if the 3 least significant place values of x are one of {000000, 000008, 000016, 000024, 000032, 00...., 000368} (250047 combinations). (Regular test, richness = 3).
    Alternatively, if the third-to-last place value is even and the last 2 place values are one of {0000, 0008, 0016, 0024, 0032, ...., 5872},
    or if the third-to-last place value is odd and the last 2 place values are one of {0004, 0012, 0020, 0028, 0036, ...., 5868}, then x is divisible by 8.
  • 9 if the least significant place value of x is one of {0, 9, 18, 27, 36, …, 117} (14 combinations). (Regular test, richness = 1: divisor).
  • 10 if x is divisible by 2 and 5. (Composite test: omega inheritor).
  • 12 if the 2 least significant place values of x are one of {0000, 0012, 0024, 0036, 0048, ...., 5864} (1323 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is even and the last place value is one of {00, 12, 24, 36, 48, 60, 72, 84, 96, 08, 20},
    or if the penultimate place value is odd and the last place value is one of {06, 18, 30, 42, 54, 66, 78, 90, 02, 14}, then x is divisible by 12.
  • 14 if the least significant place value of x is one of {0, 14, 28, 42, 56, 70, 84, 98, 112}. (Regular test, richness = 1: divisor).
  • 15 if x is divisible by 3 and 5. (Composite test: omega inheritor).
  • 16 if the 4 least significant place values of x are one of {00000000, 00000016, 00000032, 00000048, 00000064, 0000...., 52047360} (15752961 combinations). (Regular test, richness = 4: impractical).
  • 18 if the least significant place value of x is one of {0, 18, 36, 54, 72, 90, 108}. (Regular test, richness = 1: divisor).
  • 20 if x is divisible by 4 and 5. (Composite test: omega inheritor).
  • 21 if the least significant place value of x is one of {0, 21, 42, 63, 84, 105}. (Regular test, richness = 1: divisor).
  • 24 if the 3 least significant place values of x are one of {000000, 000024, 000048, 000072, 000096, 00...., 000352} (83349 combinations). (Regular test, richness = 3).
    Alternatively, if the third-to-last place value is even and the last 2 place values are one of {0000, 0024, 0048, 0072, 0096, ...., 5864},
    or if the third-to-last place value is odd and the last 2 place values are one of {0012, 0036, 0060, 0084, 0108, ...., 5852}, then x is divisible by 24.
  • 25 if the centohexavigesimal digital root of x is divisible by 25. (Neighbor factor test: omega).
  • 27 if the 2 least significant place values of x are one of {0000, 0027, 0054, 0081, 0108, ...., 5849} (588 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is trine and the last place value is one of {00, 27, 54, 81, 08},
    or if the penultimate place value is overtrine and the last place value is one of {09, 36, 63, 90, 17},
    or if the penultimate place value is undertrine and the last place value is one of {18, 45, 72, 99}, then x is divisible by 27.
  • 28 if the 2 least significant place values of x are one of {0000, 0028, 0056, 0084, 0112, ...., 5848} (567 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is even and the last place value is one of {00, 28, 56, 84, 12},
    or if the penultimate place value is odd and the last place value is one of {14, 42, 70, 98}, then x is divisible by 28.
  • 30 if x is divisible by 2, 3 and 5. (Composite test: omega inheritor).


(LOTS of impractical there. Even the evenness test might be impractical! The only "practical" tests might be those associated with nonregular numbers!)

Here is dozenal:

An arbitrary integer x is divisible by …

  • 2 if the least significant place value of x is one of {0, 2, 4, 6, 8, a}. (Regular test, richness = 1: divisor).
  • 3 if the least significant place value of x is one of {0, 3, 6, 9}. (Regular test, richness = 1: divisor).
  • 4 if the least significant place value of x is one of {0, 4, 8}. (Regular test, richness = 1: divisor).
  • Coprime: Alpha-2 Factor
  • 6 if the least significant place value of x is one of {0, 6}. (Regular test, richness = 1: divisor).
  • 8 if the 2 least significant place values of x are one of {00, 08, 14, 20, 28, 34, 40, 48, 54, 60, 68, 74, 80, 88, 94, a0, a8, b4} (18 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is even and the last place value is one of {0, 8},
    or if the penultimate place value is odd and the last place value is one of {4}, then x is divisible by 8.
  • 9 if the 2 least significant place values of x are one of {00, 09, 16, 23, 30, 39, 46, 53, 60, 69, 76, 83, 90, 99, a6, b3} (16 combinations). (Regular test, richness = 2).
    Alternatively, if the penultimate place value is trine (one of {0, 3, 6, 9}) and the last place value is one of {0, 9},
    or if the penultimate place value is overtrine (one of {1, 4, 7, a}) and the last place value is one of {6},
    or if the penultimate place value is undertrine (one of {2, 5, 8, b}) and the last place value is one of {3}, then x is divisible by 9.
  • a if x is divisible by 2 and 5. (Composite test: alpha-2 inheritor).
  • b if the duodecimal digital root of x is divisible by b. (Neighbor factor test: omega).
  • 10 if the least significant place value of x is {0}. (Regular test, richness = 1: base power).
  • 11 if the alternating duodecimal sum of the place values of x is divisible by 11. (Neighbor factor test: alpha).
  • 13 if x is divisible by 3 and 5. (Composite test: alpha-2 inheritor).
  • 14 if the 2 least significant place values of x are one of {00, 14, 28, 40, 54, 68, 80, 94, a8} (9 combinations). (Regular test, richness = 2).{0}
  • 16 if the 2 least significant place values of x are one of {00, 16, 30, 46, 60, 76, 90, a6} (8 combinations). (Regular test, richness = 2).{0}
  • 18 if x is divisible by 4 and 5. (Composite test: alpha-2 inheritor).
  • 1a if x is divisible by 2 and 11. (Composite test: omega inheritor).
  • 20 if the 2 least significant place values of x are one of {00, 20, 40, 60, 80, a0} (6 combinations). (Regular test, richness = 2).{0}
  • 22 if x is divisible by 2 and 13. (Composite test: alpha inheritor).
  • 23 if the 3 least significant place values of x are one of {000, 023, 046, 069, 090, ..., b99} (64 combinations). (Regular test, richness = 3).
    Alternatively, if the third-to-last place value is trine (one of {0, 3, 6, 9}) and the last place value is one of {00, 23, 46, 69, 90, b3},
    or if the third-to-last place value is overtrine (one of {1, 4, 7, a}) and the last place value is one of {16, 39, 60, 83, a6},
    or if the third-to-last place value is undertrine (one of {2, 5, 8, b}) and the last place value is one of {09, 30, 53, 76, 99}, then x is divisible by 27.
  • 26 if x is divisible by 2, 3 and 5. (Composite test: alpha-2 inheritor).

(5 glitched. Easy fix.)

What can be more in-depth than has been done here is a description of the semitotative landscape and left/right trim tricks that are "non-intuitive". Semitotatives have been sorted out according to relatedness (coprime factor) and richness (regular factor). Decimal 6, for example, is 2 * 3; it has richness 1 and is omega related. Dozenal ten has richness 1 and alpha-2 related, etc.

At some point the tour has to be on its own website. It really sucks a lot of memory here. If it goes there then I will have to credit you for the work you and Oschkar have done. (He is already credited in the base-naming function).

As I wrote the tour I have become thoroughly convinced it has to be fully automated. There is just too high of a chance of error if we try to hand-write tables; that is the big problem. Code-generated glitches can be nipped in the bud and corrected for all output but human error or oversight leads to greater, less trackable error. The digit maps are fully automated and that eliminates the necessity to tease out whoopsies. In a similar vein prose can be generated as well. (The summary script I just wrote is sort of a tide-over so that we get forward motion: it can't handle low bases, primes, prime powers, etc., which you are not looking at in the extension).
Top
Double sharp
Posted: Nov 8 2017, 02:59 PM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



Sure, I think I've finally done and suggested enough to be comfortable with getting credited. It's just that I don't really like my real name to be shown to everyone online, so feel free to PM or email me when the site is closer to going live and I will gladly give it to you (just not post it here). happy.gif

I think the impracticality of tests for a base like 126 is a bit of a double question masquerading as a single one. It is absolutely practical if you code it as 7-on-18 or 9-on-14 and absolutely impractical as a pure base. And I am not sure the nonregular tests are as good because there are still 25 multiples of 5 to memorise. I think we can all agree that the test for 125 is practical, but then so is the test for 126; you still look for a trailing zero.

How's the website going to be like? I was thinking that most bases are not actually all that interesting, so we could probably just have the reader type in some number and an automatic chart would come up. But which bases to treat this way?

In the meantime, I think the OP should have 126 and 144 swapped (since 126 is smaller), and I think there's a typo in the title for 168. The newly completed 252 is also missing. Though I wonder if it might not be better to wait until more grand bases are done and fill them all in one fell swoop instead.

P.S. Once the base passes 120, perhaps we could start saying not what auxiliaries the base could use, but what bases could use it as an auxiliary? happy.gif I plan to get to 540 tomorrow and it strikes me that it is a really good angular auxiliary for octodecimal, where it is 1c0{i}. A shame it can only be divided by two twice; perhaps octodecimalists might be more comfortable with the admittedly significantly larger 1080 = 360{i} (lulz!). It really does seem like "360" is always a useful number in the even bases from eight to eighteen.

P.P.S. Looks like this thread has just surpassed a myriad views, which though decimal is pretty cool!
Top
Double sharp
Posted: Nov 18 2017, 04:45 PM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



Reminder to myself and to the others that the future tours are coming, but I'll have to give a few more of the huge bases a test-drive before I write something. I have 216, 336, and 540 in the works and should be able to post something useful about them shortly.

EDIT: Remaining ones are 150, 160, 192, 216, 288, 300, 336, 432, 540, 600, 630, 660.
Top
icarus
Posted: Nov 19 2017, 09:46 PM


Dozens Demigod


Group: Admin
Posts: 1,913
Member No.: 50
Joined: 11-April 06



Code writing has resumed, see this post for the latest. Most of the code work regards tapping the register for data and enveloping it in prose that doesn't sound like robospeak. A tremendous amount of the writing can be automated, and I think (as said many times before) it ought to be, since most of the properties are number-theoretical and easily calculated. The tedium and magnitude of writing an entry, especially tables of dozens or hundreds of colored cells, is a job for the computer and not a poor soul.

It looks like we'll lose HTML capabilities shortly. (Please don't respond to this here, try this thread.) Because of this I am stepping up the web version of the Tour. In a pinch all the code we wrote can be put up at that site.
Top
Double sharp
Posted: Nov 19 2017, 11:42 PM


Dozens Disciple


Group: Members
Posts: 1,402
Member No.: 1,150
Joined: 19-September 15



We are also copying over the text that we already have, right? It is pure HTML and there shouldn't be any problem putting those existing entries up with the tables in a pinch. The important thing is to archive all the stuff first. (I mostly put up 50 because I already had it mostly done, but truth be told it's not so interesting; it's like 40 but even clunkier. It just illustrates how the pattern of 18 looks in the heights.)

For a few low bases like 14, 16, 20, 21, and 35, there has been an immense amount of conversation after the main tour posts, and there may be some extra tables past the first page as well that we need to grab. But then again, many of these can be automatically generated. For example, I would like to finally have full multiplication tables for all bases up to 30 and 36 inclusive!! ^_-☆

(The hexadecimal post is an interesting one. It covers the possible civilisational use and would seem to suggest some basic ideas about classifying bases and concerns about civilisationality to be put up too. So I need to gather together a few threads and sort out the material I've written.)
Top
« Next Oldest | Non-Dozenal bases | Next Newest »
DealsFor.me - The best sales, coupons, and discounts for you

Topic OptionsPages: (10) « First ... 8 9 [10] 



Hosted for free by zIFBoards* (Terms of Use: Updated 2/10/2010) | Powered by Invision Power Board v1.3 Final © 2003 IPS, Inc.
Page creation time: 0.0412 seconds · Archive