What is the Collatz Problem and how can it be solved?

  • Thread starter Organic
  • Start date
In summary, the conversation is about the Collatz problem and a paper that proposes a proof for it. The conversation highlights issues with the clarity and correctness of the proof, including unclear language and incorrect use of mathematical symbols. The paper's author defends their work and explains their unconventional approach to the problem, but ultimately it is pointed out that their proof does not actually prove anything significant.
  • #386
Nope, there's still no reason at all to conclude that the list is in anyway a set of card 2^aleph-0, there is no reason to suppose it contains all the elements of the set of all infinite strings of 0s and 1s, indeed the list STILL contains only those strings that have a finite number of 0s on them as has been proven to you. Your only proof is that it can be no other thing... erm, not true. As I've asked, and Hurkyl, where is the string ..1010101 of alternating 0s and 1s?

So, I've read the article AGAIN. WHy don't you explain where the counter proofs of you assertions are wrong in your opinion. Remember when we asked how to construct the diagram? And we agreed the th first column is (1010101010... ) the second (110011001100..) and so on - the nth is 2^n ones, 2^n 0s, looping again and again?

remember how we showed you that that implies that every row has only a finite number of zeroes in it? remember how that implies the string ...01010101) with an infinite number of 0s in it is not on the list? remember? come on, we;ve read the article, we've said what we consider wrong with it, and it is encapsulated in this paragraph and the previous one. so where are we wrong. come on, explain it in clear simple words for us that can't understand your maths, tell us where we 've gone wrong in the analysis of the diagram.
 
Physics news on Phys.org
  • #387
Matt,

...01010101 or ...10101010 is in the list, for example:

Let us take again our set:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
[b]{[/b]...,1,1,1,1[b]}[/b]<--> 1
 ...,1,1,1,0 <--> 2
 ...,1,1,0,1 <--> 3 
 ...,1,1,0,0 <--> 4 
 ...,1,0,1,1 <--> 5 
 ...,1,0,1,0 <--> 6 
 ...,1,0,0,1 <--> 7 
 ...,1,0,0,0 <--> 8 
 ...,0,1,1,1 <--> 9 
 ...,0,1,1,0 <--> 10
 ...,0,1,0,1 <--> 11
 ...,0,1,0,0 <--> 12
 ...,0,0,1,1 <--> 13
 ...,0,0,1,0 <--> 14
 ...,0,0,0,1 <--> 15
 ...,0,0,0,0 <--> 16
 ...
Now let us make a little redundancy diet:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
...  [b]1[/b]-1-1-1 <--> 1
     \  \ \0 <--> 2
      \  0-1 <--> 3 
       \  \0 <--> 4 
       [b]0[/b]-[b]1[/b]-1 <--> 5 
        \ \[b]0[/b] <--> 6 
         0-1 <--> 7 
          \0 <--> 8 
 ... [b]0[/b]-[b]1[/b]-1-1 <--> 9 
     \  \ \0 <--> 10
      \  [b]0[/b]-[b]1[/b] <--> 11
       \  \0 <--> 12
       0-1-1 <--> 13
        \ \0 <--> 14
         0-1 <--> 15
          \0 <--> 16
 ...
and we get:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
          /1 <--> 1
         1 
        / \0 <--> 2
       1   
       /\ /1 <--> 3 
      /  0
     /    \0 <--> 4 
 ... [b]1[/b]    
     \    /1 <--> 5 
      \  [b]1[/b] 
       \/ \[b]0[/b] <--> 6
       [b]0[/b]  
        \ /1 <--> 7
         0
          \0 <--> 8
          
          /1 <--> 9 
         1
        / \0 <--> 10
       [b]1[/b]  
       /\ /[b]1[/b] <--> 11
      /  [b]0[/b] 
     /    \0 <--> 12
 ... [b]0[/b]    
     \    /1 <--> 13
      \  1
       \/ \0 <--> 14
       0  
        \ /1 <--> 15
         0
          \0 <--> 16
 ...
 
Last edited:
  • #388
That isn't one of the rows though is it? It is the set of rows that needs to have cardinality both aleph-0 and 2^aleph-0, not the number of ways of choosing entries from the rows, or the number of paths through the rows. Please answer this question.

Is the description of the construction of the diagram I gave accurate? The one I gave two posts back, the one I've given several times.

YES or NO? Can't say fairer than that; all we want in your next post is exactly on word, yes, or no, which is it?
 
Last edited:
  • #389
Matt,
That isn't one of the rows though is it? It is the set of rows that needs to have cardinality both aleph-0 and 2^aleph-0, not the number of ways of choosing entries from the rows, or the number of paths through the rows.
My (aleph0 x 2^aleph0) matrix and an Infinitely (Width x Length) Binary Tree are two representations of the same thing.
 
  • #390
Is the description of the construction of the diagram I gave accurate? The one I gave FOUR posts back, the one I've given several times.

YES or NO? Can't say fairer than that; all we want in your next post is exactly on word, yes, or no, which is it?
 
  • #391
Matt,
1) And we agreed the th first column is (1010101010... ) the second (110011001100..) and so on - the nth is 2^n ones, 2^n 0s, looping again and again?

2) remember how we showed you that that implies that every row has only a finite number of zeroes in it? remember how that implies the string ...01010101) with an infinite number of 0s in it is not on the list?

1) Yes I agree that this is the redundant way to show how the columns are constructed, but:

2) You showed nothing because the three representations below are one and only one thing:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
[b]{[/b]...,1,1,1,1[b]}[/b]<--> 1
 ...,1,1,1,0 <--> 2
 ...,1,1,0,1 <--> 3 
 ...,1,1,0,0 <--> 4 
 ...,1,0,1,1 <--> 5 
 ...,1,0,1,0 <--> 6 
 ...,1,0,0,1 <--> 7 
 ...,1,0,0,0 <--> 8 
 ...,0,1,1,1 <--> 9 
 ...,0,1,1,0 <--> 10
 ...,0,1,0,1 <--> 11
 ...,0,1,0,0 <--> 12
 ...,0,0,1,1 <--> 13
 ...,0,0,1,0 <--> 14
 ...,0,0,0,1 <--> 15
 ...,0,0,0,0 <--> 16
 ...
Now let us make a little redundancy diet:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
...  [b]1[/b]-1-1-1 <--> 1
     \  \ \0 <--> 2
      \  0-1 <--> 3 
       \  \0 <--> 4 
       [b]0[/b]-[b]1[/b]-1 <--> 5 
        \ \[b]0[/b] <--> 6 
         0-1 <--> 7 
          \0 <--> 8 
 ... [b]0[/b]-[b]1[/b]-1-1 <--> 9 
     \  \ \0 <--> 10
      \  [b]0[/b]-[b]1[/b] <--> 11
       \  \0 <--> 12
       0-1-1 <--> 13
        \ \0 <--> 14
         0-1 <--> 15
          \0 <--> 16
 ...
and we get:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
          /1 <--> 1
         1 
        / \0 <--> 2
       1   
       /\ /1 <--> 3 
      /  0
     /    \0 <--> 4 
 ... [b]1[/b]    
     \    /1 <--> 5 
      \  [b]1[/b] 
       \/ \[b]0[/b] <--> 6
       [b]0[/b]  
        \ /1 <--> 7
         0
          \0 <--> 8
          
          /1 <--> 9 
         1
        / \0 <--> 10
       [b]1[/b]  
       /\ /[b]1[/b] <--> 11
      /  [b]0[/b] 
     /    \0 <--> 12
 ... [b]0[/b]    
     \    /1 <--> 13
      \  1
       \/ \0 <--> 14
       0  
        \ /1 <--> 15
         0
          \0 <--> 16
 ...
 
Last edited:
  • #392
So we have absolutely nailed the construction of the array.

Now, are you claiming the the cardinality of the rows is 2^aleph-0 because for every element in the power set of N, there is a row which corresponds to the indicator function of that element in the power set? Yes, or No is again all that is required. Note this is equivlaent to saying that the set of rows has card 2^aleph-0 because the rows form a 'complete' list of every string of 0s and 1s, ie *ANY* possible string of 0s and 1s occurs as one of the rows.
 
  • #393
Matt,

If we use your terminology, then the answer is yes.
 
  • #394
But we agreed that we are using the cantor view point on cardinality, didn't we?

So, let z be some arbitrary element in P(N), it must if your conjecture is true be one of the rows. Let the number of the row (they're enumerated by N) be R.

Are we ok so far? Yes or no?
 
  • #395
Matt,
(they're enumerated by N)
No, they're enumerated.

This is what you don't understand, the list of unique notations that related to R is longer then the list of unique notations that related to N.

In your termenology the complete R list is longer than the complete N list.
 
Last edited:
  • #396
So the rows are not of cardinality aleph-0 as you've been claiming all through this, then?
 
  • #397
Matt,

You simply don't read what I write.

So here it is again:

1) Both R list and N list are enumerable.

2) |R|>|N|
 
  • #398
So, when I asked you if the rows could be enumerated by N, and you said no, were you lying or mistaken? because that's what enumerable means in this context. so we ask again, are the set of rows in bijective correspondence with N as you claim they are? So picking an element z in P(N) it occurs at row r for some r in N (this r was the R last time, i didn't mean R as in real number, sorry if that's the confusion), that is what you mean by the rows are enumerable, which they are by the construction we've agreed on. And you are claiming that each element of P(N) corresponds to some row.
 
  • #399
Incidentally, the word enumerable is the same as countable. So, remembering that we are usgin real mathematics here, your last post states there is a bijection between R and N, and that there isn't a bijection between R and N.
 
  • #400
Matt,
are the set of rows in bijective correspondence with N as you claim they are?
You know what? let us play your game.

1) I constructed a list with unique sequences of 0 1 combinations each, which its length has 2^aleph0 magnitude.

2) I also showed a bijection from N to P(N).


Conclusion:

The transfinite univereses do not exist.
 
Last edited:
  • #401
But we are playing my game, we agreed to use the proper definitions of all the words.

So there is a bijectin from N to P(N) because there is a row in your diagram for each element of the power set - the element being given by the indicator function described by each row. that is the bijection isn't it?
 
  • #402
Matt,

We have found that |P(N)|>=|N|

Therefore the tansfinite cardinality does not exist.
 
  • #403
Please answer my question as to how the bijection you are claiming between N and P(N) arises.

here it is


So [in your opinion] there is a bijection from N to P(N) because there is a row in your diagram for each element of the power set - the element being given by the indicator function described by each row. that is the bijection isn't it?

again all that is required is a simple yes or no answer, no more.
 
  • #404
Matt,

There is no YES/NO answer here.

Conclusions:


1) |R| > |N| (R list is longer then N list).

2) There is no necessary connection between N and being enumerable.

3) The words “complete” or “all” are meaningless when they are related to a collection of infinitely many elements, because they are leading to contradiction (|R| >= |N|).

Instead, the word "any is used.

Maybe this will help:

Code:
<--------------------Arithmetic magnitude

 {...,3,2,1,0} = Z*
     2 2 2 2
     ^ ^ ^ ^         Geometric magnitude
     | | | |                  | 
     v v v v                  |    
{...,1,1,1,[b]1[/b]} <--> 1          | 
 ...,1,1,[b]1[/b],                   |
 ...,1,[b]1[/b] <---- Cantor’s       |
 ...,[b]1[/b],1, ,    Diagonal       |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...                          |
 ...,0,0,0,0 <--> 2           | 
 ...,0,0,0, ----.             |
 ...,0,0, ,     |             |
 ...,0,0, ,     | Not Covered |
 ...,0, , ,     |--  By N     |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ... -----------‘             |
 ...,1,1,1,1 <--> 3           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
...                           |
 ...,0,0,0,0 <--> 4           | 
 ...,0,0,0,                   |
 ...,0,0, ,                   |
 ...,0,0, ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...                          |
 ...,1,1,1,1 <--> 5           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
                              |
 ...                          V
 
Last edited:
  • #405
I would like to remind you that you are operating in the world of conventional mathematics and therefore all of the reasons you just gave are irrelevant, not to say misleading and erroneous.

You have stated there is a bijection between the set of rows and P(N), yet you cannot give this bijection or offer an existence proof. As your claims are supposedly constructive this is obviously a big error on your part. I can predict you will accuse me of just playing with words and misunderstanding the 'true' picture of what maths is, but I think we've seen enough of your position to see it is indefensible; you certainly haven't been able to defend it.

In particular enumerable and countable are defined by bijections with the set of Natural numbers, irrespective of what your opinion is on the matter.

Just because one set appears 'bigger than another' is not important. You think that |Q| equals |N| in whatever meaning you give to those symbols, yet clearly one is 'smaller' than the other, in the sense that a list of Q elements contains a list of N elements as a sublist, to put it in your terms.

After all this do you think that you perhaps ought to learn the meaning of any of the terms you use, because you evidently haven't even after somewhere in the region of 600 posts on it to my knowledge.
 
  • #406
Matt,

There exist a countable list of all(=your termenology) R members represented by infinitely long combinations of 01 sequences, in front of your eyes.

Prove that this list does not exist, but before that take a look again on:
Code:
<--------------------Arithmetic magnitude

 {...,3,2,1,0} = Z*
     2 2 2 2
     ^ ^ ^ ^         Geometric magnitude
     | | | |                  | 
     v v v v                  |    
{...,1,1,1,[b]1[/b]} <--> 1          | 
 ...,1,1,[b]1[/b],                   |
 ...,1,[b]1[/b] <---- Cantor’s       |
 ...,[b]1[/b],1, ,    Diagonal       |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...                          |
 ...,0,0,0,0 <--> 2           | 
 ...,0,0,0, ----.             |
 ...,0,0, ,     |             |
 ...,0,0, ,     | Not Covered |
 ...,0, , ,     |--  By [b]N[/b]     |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ... -----------‘             |
 ...,1,1,1,1 <--> 3           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
...                           |
 ...,0,0,0,0 <--> 4           | 
 ...,0,0,0,                   |
 ...,0,0, ,                   |
 ...,0,0, ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...                          |
 ...,1,1,1,1 <--> 5           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
                              |
 ...                          V
As you can see, Cantor's diagonal does not cover this list and also
N members do not cover it, therefore there is no bijection
from |N| to |P(N)| (|P(N)|>|N|).

Therefore the "bijection" that you see between R and N is only an illusion, because each
n <--> ...000/...111 capture only one slice form each 0 or 1 bulk of geometric magnitude, as you can clearly see below:
Code:
 <---Arithmetic magnitude 

 {...,3,2,1,0} = Z*
     2 2 2 2  
     ^ ^ ^ ^   
     | | | |   
     v v v v  
{...,[b]1-1-1-1[/b]} <--> 1  Geometric magnitude(based on the 
 ...,1,1,1,[b]0[/b]  <--> 2          |          thin notations)          
 ...,1,1,[b]0[/b]/                   |
 ...,1,1/0,                   |
 ...,1,[b]0[/b], ,                   |
 ...,1/0, ,                   |
 ...,1|0, ,                   |
 ...,1|0, ,                   |
 ...,[b]0[/b]/ , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...                          V

As for Q and N, both have an arithmetic magnitude.
 
Last edited:
  • #407
So now you admit there is no bijection from N to P(N)? So why did you claim that aleph-0 equals 2^aleph-0? Have you completely recanted your assertion, or are you going to claim that you've never insisted that they are the same? becaue you've repeatedly insisted that the reals are countable that implies there is a bijection from N to P(N).

Or are you going to simultaneously claim that there is a set of cardinality 2^aleph-0 that is countable, yet not in bijection with N, contradicting the definition of countable?
 
  • #408
Matt,


I said that there is a bijection from |N| to |P(N)| when I play your game, and by your game the conclusion is that the transfinite cardinality cannot exist because we get |N|=<|P(N)|.

But if we want to understand what is going on, we have no choice but to say bye bye to your game and move to another game, where R is enumerable but |R|>|N|.

This result, which is impossible trough your game, can be simply and clearly explained and demonstrated by my game.


Now we are in my game, so this time please read all that is below and try to understand it:

There exist a countable list of all(=your terminology) R members represented by infinitely long combinations of 01 sequences, in front of your eyes.

Prove that this list does not exist, but before that take a look again on:
Code:
<--------------------Arithmetic magnitude

 {...,3,2,1,0} = Z*
     2 2 2 2
     ^ ^ ^ ^         Geometric magnitude
     | | | |                  | 
     v v v v                  |    
{...,1,1,1,[b]1[/b]} <--> 1          | 
 ...,1,1,[b]1[/b],                   |
 ...,1,[b]1[/b] <---- Cantor’s       |
 ...,[b]1[/b],1, ,    Diagonal       |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...                          |
 ...,0,0,0,0 <--> 2           | 
 ...,0,0,0, ----.             |
 ...,0,0, ,     |             |
 ...,0,0, ,     | Not Covered |
 ...,0, , ,     |--  By [b]N[/b]     |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ...,0, , ,     |             |
 ... -----------‘             |
 ...,1,1,1,1 <--> 3           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
...                           |
 ...,0,0,0,0 <--> 4           | 
 ...,0,0,0,                   |
 ...,0,0, ,                   |
 ...,0,0, ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...                          |
 ...,1,1,1,1 <--> 5           | 
 ...,1,1,1,                   |
 ...,1,1, ,                   |
 ...,1,1, ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
 ...,1, , ,                   |
                              |
 ...                          V
As you can see, Cantor's diagonal does not cover this list and also
N members do not cover it, therefore there is no bijection
from |N| to |P(N)| (|P(N)|>|N|).

Therefore the "bijection" that you see between R and N is only an illusion, because each
n <--> ...000/...111 capture only one slice form each 0 or 1 bulk of geometric magnitude, as you can clearly see below:
Code:
 <---Arithmetic magnitude 

 {...,3,2,1,0} = Z*
     2 2 2 2  
     ^ ^ ^ ^   
     | | | |   
     v v v v  
{...,[b]1-1-1-1[/b]} <--> 1  Geometric magnitude(based on the 
 ...,1,1,1,[b]0[/b]  <--> 2          |          thin notations)          
 ...,1,1,[b]0[/b]/                   |
 ...,1,1/0,                   |
 ...,1,[b]0[/b], ,                   |
 ...,1/0, ,                   |
 ...,1|0, ,                   |
 ...,1|0, ,                   |
 ...,[b]0[/b]/ , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...,0, , ,                   |
 ...                          V

As for Q and N, both have an arithmetic magnitude.



Conclusions:


1) |R| > |N| (R list is longer then N list).

2) There is no necessary connection between N and being enumerable.

3) The words “complete” or “all” are meaningless when they are related to a collection of infinitely many elements, because they are leading to contradiction (|R| >= |N|).

Instead, the word "any" is used.
 
Last edited:
  • #409
I am perfectly open to new ideas, always have been, however, you have repeatedly said that proper mathematics is wrong and its error means that Cantor is wrong, and that there is only one kind of infinity. It doesn't matter what your definitions and issues are here, only that you must then prove something is wrong in Cantor's argument. Your method is to then redefine all of the terms to mean something else. Thus you are being a complete moron in claiming he is wrong, because you are deliberately misinterpreting what he said.


I don't particularly care for your opinion on your new mathematics, I don't see anything interesting in it, nor new. My only position has been that you do not understand tht which you defame and that we cane easliy prove Cantor is correct because you evidently don't have a clue what's going on. Shall we agree on that? I don't care about your opinion, and you know squat about mathematics?

There is no bijection in my mathematical world from N to P(N). So why do you keep saying that? It's always been your contention that there must be, but you've never been able to prove it, and now you're claiming you never wanted to...? Pillock.
 
  • #410
Matt,

Ok, prove by your system that my matrix does not have the complete 01 combinations.

...0101 or ...1010 is in the list, for example:

Let us take again our set:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
[b]{[/b]...,1,1,1,1[b]}[/b]<--> 1
 ...,1,1,1,0 <--> 2
 ...,1,1,0,1 <--> 3 
 ...,1,1,0,0 <--> 4 
 ...,1,0,1,1 <--> 5 
 ...,1,0,1,0 <--> 6 
 ...,1,0,0,1 <--> 7 
 ...,1,0,0,0 <--> 8 
 ...,0,1,1,1 <--> 9 
 ...,0,1,1,0 <--> 10
 ...,0,1,0,1 <--> 11
 ...,0,1,0,0 <--> 12
 ...,0,0,1,1 <--> 13
 ...,0,0,1,0 <--> 14
 ...,0,0,0,1 <--> 15
 ...,0,0,0,0 <--> 16
 ...
Now let us make a little redundancy diet:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
...  [b]1[/b]-1-1-1 <--> 1
     \  \ \0 <--> 2
      \  0-1 <--> 3 
       \  \0 <--> 4 
       [b]0[/b]-[b]1[/b]-1 <--> 5 
        \ \[b]0[/b] <--> 6 
         0-1 <--> 7 
          \0 <--> 8 
 ... [b]0[/b]-[b]1[/b]-1-1 <--> 9 
     \  \ \0 <--> 10
      \  [b]0[/b]-[b]1[/b] <--> 11
       \  \0 <--> 12
       0-1-1 <--> 13
        \ \0 <--> 14
         0-1 <--> 15
          \0 <--> 16
 ...
and we get:
Code:
 {...,3,2,1,0}=Z*
     2 2 2 2
     ^ ^ ^ ^
     | | | |
     v v v v
          /1 <--> 1
         1 
        / \0 <--> 2
       1   
       /\ /1 <--> 3 
      /  0
     /    \0 <--> 4 
 ... [b]1[/b]    
     \    /1 <--> 5 
      \  [b]1[/b] 
       \/ \[b]0[/b] <--> 6
       [b]0[/b]  
        \ /1 <--> 7
         0
          \0 <--> 8
          
          /1 <--> 9 
         1
        / \0 <--> 10
       [b]1[/b]  
       /\ /[b]1[/b] <--> 11
      /  [b]0[/b] 
     /    \0 <--> 12
 ... [b]0[/b]    
     \    /1 <--> 13
      \  1
       \/ \0 <--> 14
       0  
        \ /1 <--> 15
         0
          \0 <--> 16
 ...
you have repeatedly said that proper mathematics is wrong and its error means that Cantor is wrong, and that there is only one kind of infinity.
My system is reacher then Cantor's transfinite universes bacause:

1) By my system aleph0+1 > aleph0 , 2^aleph0 < 3^aleph0

2) By Cantor's system aleph0+1 = aleph0 , 2^aleph0 = 3^aleph0


By the way, when we move from the 01 matrix representation to the Binary Tree representation, the meaning of the word magnitude become clearer, because several sequential 1 or 0 notations of each column in the matrix, are compressed to a single notation, which its magnitude equivalent to the quantity of the notations that it represents.
 
Last edited:
  • #411
My new motto will be don't feed the trolls.

My new motto will be don't feed the trolls.


sory i did not answer you matt
i am to busy

but way you call Organic a Troll ?

Don't you see he is just a monkey ...

And aren't we all the monkey of Euclid ?


Moshek
 
Back
Top