« first day (2594 days earlier)      last day (2724 days later) » 

08:01
what programming language is this?
Prover9 is an automated theorem prover for First-order and equational logic developed by William McCune. Prover9 is the successor of the Otter theorem prover. Prover9 is intentionally paired with Mace4, which searches for finite models and counterexamples. Both can be run simultaneously from the same input, with Prover9 attempting to find a proof, while Mace4 attempts to find a (disproving) counter-example. Prover9, Mace4, and many other tools are built on an underlying library named LADR to simplify implementation. Resulting proofs can be double-checked by Ivy, a proof-checking tool that has been...
It managed to prove the existence of left inverse from the existence of right inverse + identity in 0.03 seconds
Assumptions:
all a exists b (a*b = 1).
all a all b all c a*(b*c) = (a*b)*c.
all a a*1 = a.
Goals:
all x exists y (y*x = 1).
Proof:
1 (all a exists b a * b = 1) # label(non_clause).  [assumption].
2 (all a all b all c a * (b * c) = (a * b) * c) # label(non_clause).  [assumption].
3 (all a a * 1 = a) # label(non_clause).  [assumption].
4 (all x exists y y * x = 1) # label(non_clause) # label(goal).  [goal].
5 x * f1(x) = 1.  [clausify(1)].
6 (x * y) * z = x * (y * z).  [clausify(2)].
7 x * 1 = x.  [clausify(3)].
8 x * c1 != 1.  [deny(4)].
10 x * (f1(x) * y) = 1 * y.  [para(5(a,1),6(a,1,1)),flip(a)].
11 x * (1 * y) = x * y.  [para(7(a,1),6(a,1,1)),flip(a)].
Hmm, I wonder what it will return when I put in the division by zero algebra axioms, hopefully it is robust enough to handle an arbitrary ringnoid
I already know the result for the finite case, but the infinite case is still open for associative algebras
(and I am too busy at chemistry to do further investigations in maths in general)
@Secret give me your axioms and your goals
I'll help you type them in
Axioms for infinite associative division by zero algebra:
1. $|S|$ is infinite
2. left additive identity 0+x=x for all x.
3. right zero inverse 0q=1
4. (I forgot whether left or right) multiplicative identity 1
5. Left distributive law a(b+c)=ab+ac
6. Associative law

(I might have mixed up some of the left right combinations, as the master documents is not with me currently)
I can't code 1
(and it's assumed implicitly anyway)
08:08
ok then put in 2-6 and see what the program said about it
@Secret oh, we're just checking if it's consistent?
Yes, if the No-Go Theorem is true for all associative algebras, I expect associative law will break
I already know the case for finite sets is relatively trivial with null semigroup stuff
If I do recall all axioms correctly, the program should not be able to simplify any products of zeros
and will just output them as it is
I will be back after the bath, let me know if the program found anything interesting
I still don't understand what your goal is...
 + :
       | 0 1 2
    ---+------
     0 | 0 1 2
     1 | 0 1 2
     2 | 0 1 2

 * :
       | 0 1 2
    ---+------
     0 | 2 0 1
     1 | 0 1 2
     2 | 1 2 0

 q : 2
The above is the "counter-example" generated for the assumptions below:
all x 0+x=x.
0*q=1 & q != 0 & q != 1.
all x x*1=x.
all x 1*x=x.
all a all b all c a*(b+c)=(a*b)+(a*c).
all a all b all c (a*b)*c=a*(b*c).
all a all b all c (a+b)+c=a+(b+c).
(it's actually an example for reasons I won't try to explain)
08:40
Back.
The above case is already known because all finite associative division by zero algebras must be composed of null semigroups, its extensions or a direct product of them (as it is evident in the + table that it is a right null semigroup)

What I am interested in is whether in the infinite case there exists examples that does not contain null semigroups
what is a null semigroup?
and can you check the set of axioms?
A null semigroup is a semigroup with an absorbing element. The element can be the unique two sided absorber, or can be either left or right absorbers
It's cayley table will be either a column of the same types of elements, or a row of them, or the whole table being the absorber, as shown in the wikipedia examples
so your goal is to prove that it must be a null semigroup?
What I proved last year is that all finite associative division by zero algebras must be a left or right null semigroup, or a concatenation of the cayley table of a finite number of right/left null semigroups. This is the No-Go Theorem that I always talked about

But I have yet to prove whether it also holds for infinite structures
The reasons for that is that in infinite dimensions, an injective map is not necessary surjective, thus it might provide an opening to allow "nontrivial" structures
 + :
       | 0 1 2 3
    ---+--------
     0 | 0 1 2 3
     1 | 0 1 2 3
     2 | 0 1 2 3
     3 | 3 3 3 3

 * :
       | 0 1 2 3
    ---+--------
     0 | 2 0 1 3
     1 | 0 1 2 3
     2 | 1 2 0 3
     3 | 3 3 3 3

 q : 2
Counter-example of the newly-added goal all x all y x+y=y.
08:48
No finite cases will be what I am looking for. What you generated here is the concatenation of the trivial semigroup {3} with the 3x3 right null semigroup (0,1,2)
It is still quite amazing that the program can re-generate these results so quickly while back then it took me at least 3 days to come up an example
I can't quantify over subsets in first order logic.
A corollary that I used to prove the no go theorem is that if the multiplication table contains a latin square, then its corresponding + table must be a null semigroup
@Secret would you just tell me what examples you want to see?
I want an example where you cannot find any columns/rows or subcolumns/subrows with the same element. For example, that L shape border of 3 is bad, the column of 1 and 2 is also bad, which according to the theorem, the set must be infinite.
Let me check the program to see whether it can do presentation-esch things to display infinite structures...
In order to do so, it means the * table cannot contain a latin square
(actually... I am not very sure if the latin square requirement is true for infinite sets, because when I do that proof, the fact that injective maps does no imply surjective for infinite sets allows an escape and thus I failed to prove that for the infinite case, thus it is still open)
@Secret for the addition or the multiplication?
08:55
Actually, I need BOTH tables to do that, which is why the structure must be infinite
look, I can't generate infinite examples.
I see, sorry...
 + :
       | 0 1 2 3 4
    ---+----------
     0 | 0 1 2 3 4
     1 | 0 1 2 3 4
     2 | 0 1 2 3 4
     3 | 4 4 4 3 4
     4 | 4 4 4 3 4

 * :
       | 0 1 2 3 4
    ---+----------
     0 | 2 0 1 3 4
     1 | 0 1 2 3 4
     2 | 1 2 0 3 4
     3 | 3 3 3 3 3
     4 | 4 4 4 3 4
I noticed the 3 column in the multiplication, wait
It's alright, if I have done my proof correctly, the program should not be able to generate any finite examples that can fullfill the requirement.

(I will be quite interested if it does, but then it must be a very large number cause the largest number of elements I have tested is 10 before I start writing the proof)
the program is not doing well
it will auto destruct in 1 minute
09:01
This particular 5x5 + table is quite interesting, though, I do not recall seeing something like 4 4 4 3 4 before...
time limit exceeded :D
removing one of my constraints gives you this:
 + :
       | 0 1 2 3 4
    ---+----------
     0 | 0 1 2 3 4
     1 | 0 1 2 3 4
     2 | 0 1 2 3 4
     3 | 4 4 4 3 4
     4 | 4 4 4 3 4

 * :
       | 0 1 2 3 4
    ---+----------
     0 | 2 0 1 3 4
     1 | 0 1 2 3 4
     2 | 1 2 0 3 4
     3 | 3 3 3 4 3
     4 | 4 4 4 3 4

 q : 2
hmm....
* 3 *
3 4 3
* 3 *
that seems new... I will check later and see if I have such examples...
 + :
       | 0 1 2 3 4 5 6
    ---+--------------
     0 | 0 1 2 3 4 5 6
     1 | 0 1 2 3 4 5 6
     2 | 0 1 2 3 4 5 6
     3 | 3 3 3 3 5 5 3
     4 | 4 4 4 6 4 4 6
     5 | 5 5 5 3 5 5 3
     6 | 6 6 6 6 4 4 6

 * :
       | 0 1 2 3 4 5 6
    ---+--------------
     0 | 2 0 1 3 4 5 6
     1 | 0 1 2 3 4 5 6
     2 | 1 2 0 3 4 5 6
     3 | 3 3 3 3 3 3 3
     4 | 4 4 4 4 4 4 4
     5 | 5 5 5 5 5 5 5
     6 | 6 6 6 6 6 6 6

 q : 2
ignoring multiplication gives you the above
this is so beautiful
Now my goals are:
  (all x all y x + y = y)
| (all x all y x * y = y)
| exists x (x != 0 & all y x + y = x)
| exists x (x != 0 & all y y + x = x)
| exists x (x != 0 & all y x * y = x)
| exists x (x != 0 & all y y * x = x).
I'm running the prover and the searcher at the same time
(the searcher searches for counter-examples)
@Secret do you have any specification?
both timed out.
The 7x7 table you generate fits nicely to the result of the theorem I proved (hence known), but the 5x5 table may be interesting, it does not seemed to fit the result of the theorem I proved. More investigation needed
2 mins ago, by Leaky Nun
@Secret do you have any specification?
Can you list the axioms again?
After this I'm going to explore on coding the finite ZF axioms :P
09:11
$\textbf{Axioms for finite associative division by zero algebra:}$
1. $|S|$ is finite
2. left additive identity 0+x=x for all x.
3. right zero inverse 0q=1
4. Left multiplicative identity 1
5. Left distributive law a(b+c)=ab+ac
6. Associative law
$\textbf{No-Go Theorem}$
Let $(S,+,*)$ be a division by zero algebra. If $|S| < \infty$ and associativity holds, then $(S,+,*)$ must contain at least one null semigroup like substructure in the + or * structure
@LeakyNun Don't worry, do the ZF stuff first, it's not urgent now
@Secret could you remember whether left or right?
come on...
@LeakyNun Let's say left
also, is the associative law for addition or multiplication?
Both (else it is not an associative algebra)
+ :
       | 0 1 2 3
    ---+--------
     0 | 0 1 2 3
     1 | 1 0 3 2
     2 | 1 0 3 2
     3 | 0 1 2 3

 * :
       | 0 1 2 3
    ---+--------
     0 | 3 2 1 0
     1 | 0 1 2 3
     2 | 3 2 1 0
     3 | 0 1 2 3
Searcher found.
This is most exciting.
09:14
I think I might be too strict to call
3 2 1 0
3 2 1 0
a null semigroup...
But sure
@Secret any additional rule?
27 mins ago, by Leaky Nun
I can't quantify over subsets in first order logic.
We cannot check for substructures, so I guess the only thing we can do is find some particular examples within by adding extra assumptions
Let me think...
try adding 1+3=2
After this one I think I will like to see your ZF stuff instead
@Secret look, just tell me what you want to do
it might not require quanitfying over subsets
Ok here's the recipe for this particular cases:
1. |S| is finite
2. left additive identity 0+x=x for all x.
3. right zero inverse 0q=1
4. Left multiplicative identity 1
5. Left distributive law a(b+c)=ab+ac
6. Associative law
7. 1+2=3
@Secret I haven't started yet
 + :
       | 0 1 2 3
    ---+--------
     0 | 0 1 2 3
     1 | 1 0 3 2
     2 | 1 0 3 2
     3 | 0 1 2 3

 * :
       | 0 1 2 3
    ---+--------
     0 | 3 2 1 0
     1 | 0 1 2 3
     2 | 3 2 1 0
     3 | 0 1 2 3

 q : 2
I can't directly code "1+2=3" for some reason, so I replaced it with "1+q != q"
(since q is automatically 2)
@Secret anything new?
09:26
No, it's all good for now
While I do have another algebraic structure that is not a zero term algebra I am interested in, I forgot its axioms and will check later
@Secret you expressed concern over the repeated rows
should I elimiante them?
should I make addition a latin square?
Hi I have a question regarding the sum of absolute value:
$2 \sum_{j=0}^{i} |2j-i|$ is it equal to $2 \sum_{j=0}^{i} 2j-i + 2 \sum_{j=0}^{i} -2j+i| $
@LeakyNun Do not allow latin squares, and eliminate the repeated rows. If my proof is correct, it should failed to give any examples except the trivial ring
Is this correct?!
@Secret do not allow?
09:31
It can be proved by homomorphism over + that any latin square in * will result in the + structure to be a null semigroup because given $0+x=x$, $\phi (0)+\phi (x) = \phi (x)$
thus latin squares turn multplication into a permutation map
so do not care if it is a latin square?
but you mentioned that 3 2 1 0 3 2 1 0 is bad
I mean, avoid anything that look like e.g.:
1 2 3
2 3 1
3 1 2
to appear in the tables
and remove repeated rows
time limit exceeded
there must be repeated rows.
Right, so that means my proof is ok
For some reason 0 and 1 are the only allowed numbers
if I need 2 I would have to make it q and specify that q != 0 & q != 1
09:39
I guess that's just how the program works and notate elements
^ what class is that?
@usukidoll class?
ya or what book did that info came from?
I just searched online
09:46
Hi
Could someone tell me what SO and SE mean in this text?
*SO(3) and SE(3)
@LeakyNun Thanks!
Assumptions:
all x all y all z ((x*y)*z = x*(y*z)). %G1
all x (x*e=x & x=e*x). %G2
all x exists y (x*y=e & e=y*x). %G3
Goals:
all x all y x*y=y*x.
The following table for the smallest non-abelian group is generated in 0 seconds:
   | 0 1 2 3 4 5
---+------------
 0 | 0 1 2 3 4 5
 1 | 1 0 3 2 5 4
 2 | 2 4 0 5 1 3
 3 | 3 5 1 4 0 2
 4 | 4 2 5 0 3 1
 5 | 5 3 4 1 2 0
This is very exciting.
(cc @Secret)
10:03
Something 6, let me check...
@Secret this isn't related to the zero-division thing
I know, but I forgot the smallest non abelian group's name
10:16
Assumptions:
all x all y all z ((x*y)*z = x*(y*z)). %G1
all x x*e=x. %G2
all x x*inv(x)=e. %G3
Goal:
all x inv(inv(x))=x.
Proof:
1 (all x all y all z (x * y) * z = x * (y * z)) # label(non_clause).  [assumption].
2 (all x x * e = x) # label(non_clause).  [assumption].
3 (all x x * inv(x) = e) # label(non_clause).  [assumption].
4 (all x inv(inv(x)) = x) # label(non_clause) # label(goal).  [goal].
5 (x * y) * z = x * (y * z).  [clausify(1)].
6 x * e = x.  [clausify(2)].
7 x * inv(x) = e.  [clausify(3)].
8 inv(inv(c1)) != c1.  [deny(4)].
9 x * (e * y) = x * y.  [para(6(a,1),5(a,1,1)),flip(a)].
10 x * (inv(x) * y) = e * y.  [para(7(a,1),5(a,1,1)),flip(a)].
@Secret
Right $D_6$ dihedral group of 6 elements
@Secret right, it's also $S_3$ the symmetric group of 3 letters
@AlessandroCodenotti buongiorno
@Secret is the proof legible to you?
Expanded proof:
1 (all x all y all z (x * y) * z = x * (y * z)) # label(non_clause).  [assumption].
2 (all x x * e = x) # label(non_clause).  [assumption].
3 (all x x * inv(x) = e) # label(non_clause).  [assumption].
4 (all x inv(inv(x)) = x) # label(non_clause) # label(goal).  [goal].
5 (x * y) * z = x * (y * z).  [clausify(1)].
6 x * e = x.  [clausify(2)].
7 x * inv(x) = e.  [clausify(3)].
8 inv(inv(c1)) != c1.  [deny(4)].
9A x * y = x * (e * y).  [para(6(a,1),5(a,1,1))].
9 x * (e * y) = x * y.  [copy(9A),flip(a)].
hi @Fawad
@LeakyNun hi!
> para(47(a,1),28(a,1,2,2,1)) -- paramodulate from the clause 47 into clause 28 at the positions shown.
@LeakyNun stuff in square brackets and the symbol c1 is not very understood by me, but the proof itself otherwise is legible
10:30
@Secret the stuff in square brackets is the steps used
paramodulate is basically "substitute"
% number = 1
% seconds = 0

% Interpretation of size 6

 * :
       | 0 1 2 3 4 5
    ---+------------
     0 | 0 1 2 3 4 5
     1 | 1 0 3 2 5 4
     2 | 2 4 0 5 1 3
     3 | 3 5 1 4 0 2
     4 | 4 2 5 0 3 1
     5 | 5 3 4 1 2 0

 e : 0

 c1 : 1

 c2 : 2

 inv :
         0 1 2 3 4 5
    ----------------
         0 1 2 4 3 5
it actually gives you a table for the inverse
oh, and of course c1 and c2 are where the goal fails: 1*2=3!=4=2*1
counter-example 1, I suppose
@Secret prover9 proves everything by contradiction
@Secret you might be more interested in the following puzzle:
of course prover9 has a built-in equal, but I decided to build my own
Assumptions:
all x equal(x,x).
all x all y (equal(x,y) -> equal(y,x)).
all x all y all z (equal(x,y) & equal(y,z) -> equal(x,z)).
all x equal(add(x,0),x).
all x all y equal(add(x,S(y)),S(add(x,y))).
Goals:
equal(add(S(0),S(0)),S(S(0))).
Can you construct a counter-example?
Also, why does it fail?
@Danu can I ask you a favour?
4
Q: Question regarding Gambler's Ruin

Little RookieConsider a gambling process $(X_n)_{n∈\mathbb{N}}$ on the state space $S = {0, 1, . . . , N}$, with probability $p$, resp. $q$, of moving up, resp. down, at each time step. For $x = 0, 1, . . . , N$, let $τ_x$ denote the first hitting time, $τ_x := \inf\{n ≥ 0 : X_n = x\}$ Let $p_x := P(τ_{x+1} <...

Let me translate:
\begin{align}
x & \sim x\\
x \sim y & \implies y \sim x\\
x \sim y & \text{ and } y \sim z \implies x \sim z\\
x + 0 & \sim x\\
x+S(y) & \sim S(x+y)
\end{align}
=> S(0)+S(0)=S(S(0))
10:45
you forgot the symmetric
btw, some theory: our input is in the form of "given A prove B". What prover9 does is to prove that A and not B leads to contradiction. Why it does this is because A and B can be consistent, at the same time with A and not B being consistent.
Hi, is there non-dogmatic reasoning (not based on rules, principles, axioms or doctrines) ?
@Dattier bonjour, as-tu lu ma reponse sur ta question?
Non, je ne l'ai pas vu
19 hours ago, by Leaky Nun
@Dattier $\begin{array}{cl} &\gcd(2^m+1,2^n+1) \\=& \gcd(2^m+1-2^{n-m}(2^n+1),2^n+1) \\=& \gcd(2^{n-m}+1,2^n+1) \end{array}$
19 hours ago, by Leaky Nun
and $\gcd(2^a+1,2^0+1)=2$ when $a$ is odd
la preuve de ton theoreme
Hi.
Why $\left|\dfrac{xyz}{x^2+y^2+z^2}\right| \leq \left|\dfrac{x}{x^2+y^2+z^2}\right| \leq |x| \leq \left|\sqrt{x^2+y^2+z^2}\right |$ is not true?
For all $x,y,z \in \mathbb{R}$
10:55
$$\begin{array}{cl} &\gcd(2^m+1,2^n+1) \\=& \gcd(2^m+1-2^{n-m}(2^n+1),2^n+1) \\=& \gcd(2^{n-m}+1,2^n+1) \end{array} \text{ and }\\ \gcd(2^a+1,2^0+1)=2$ when $a$ is odd$$
the first inequality fails when $|yz|>1$ @Topologicalife
Right, uhm
@LeakyNun : why you need a odd ?
sorry, $\gcd(2^a+1,2^0+1)=1$.
no matter $a$ is odd or even.
$x \leq \sqrt{x^2+y^2}$ and $y \leq \sqrt{x^2+y^2}$ so is true that $xy \leq x^2+y^2$?
11:00
But the result is not true when a, and b is even, for exemple a=4, b=2
@LeakyNun
$\gcd(2^4+1,2^2+1) = \gcd(17,5) = 1$
interesting
and 2^2+1=5
gcd(2,4)=2
$\gcd(2^m+1,2^n+1) = \gcd(2^m+1-2^{m-n}(2^n+1),2^n+1) = \gcd(-2^{m-n}+1,2^n+1)$
my proof above was wrong
method abandoned :P
11:20
S(0) + 0 = S(0) = S(0+0)
still looking...
MEanwhile S(1)+S(1)=SS(1) is wrong since 2+2=3 (assuming you are doing the usual peano arithmetic)
Actually, give only these 5 assumptions, how does one derive S(S(0))?
you don't, you just keep it as-is.
S(0)+0=S(0+0)=S(0)
SS(0)+0 = S(S(0)+0) =SS(0)
S(0)+S(0) = S(0 + S(0))=SS(0)
nope, still consistent
I suspect S(0)+S(0) then not SS(0) is inconsistent, hence the whole system is consistent
since the above workings showed that S(0)+S(0)=SS(0)
S(0)+S(0) is not a statement...
"S(0)+S(0) then not SS(0)" is meaningless
@Secret then you need to check the above workings very carefully
S(0) + 0 = S(0)
0 + S(0) = S(0+0)=S(0)
what prevents x to be an element S(0) hence writing S(0) + S(0) = S(S(0)+0)=SS(0)?
The problem is in your second line
11:32
Let x = 0, y = 0, then x + S(y) = 0 + S(0). Then use x+ S(y) = S(x+y), this gives 0 + S(0) = S(0+0) = S(0)?
@Secret there is one unjustified step here.
@privetDruzia SO(3) is the group of rotations of $\Bbb R^3$:
In mechanics and geometry, the 3D rotation group, often denoted SO(3), is the group of all rotations about the origin of three-dimensional Euclidean space R3 under the operation of composition. By definition, a rotation about the origin is a transformation that preserves the origin, Euclidean distance (so it is an isometry), and orientation (i.e. handedness of space). Every non-trivial rotation is determined by its axis of rotation (a line through the origin) and its angle of rotation. Composing two rotations results in another rotation; every rotation has a unique inverse rotation; and the identity...
SE is not really standard notation...
can we say 0 + 0 = 0 using x + 0 = x and set x=0?
@Secret yes you can
can we say x,y=0,0 and thus substitute x +S(y) to become 0 + S(0)?
11:34
@Secret yes, you can
can we then use x + S(y) = S(x+y), with x,y=0,0 to get 0 + S(0) to S(0+0)?
@Secret yes, you can
Then using:
2 mins ago, by Secret
can we say 0 + 0 = 0 using x + 0 = x and set x=0?
$S(0+0)=S(0)$, unless you mean S(0+0) and S(0) are different elements and not a map S acting on 0 and 0+0?
@Secret you can't actually prove that S(0+0)=S(0).
from 0+0=0.
that I don't understand, in terms of arguments, it makes sense, so that means S is not really a map, but something else altogether
11:38
@Secret maybe you never thought about why we can say $f(x)=f(y)$ when we know that $x=y$.
It's a property of equality that I didn't codify into my axioms.
It's actually an axiom schema.
Very likely, I just define maps directly in my algebraic structure, thus might be why I never noticed that
x=y iff P(x) iff P(y) for any proposition P
this is an axiom schema, one axiom per proposition
12:02
[Random]
x=y iff P(x) iif P(y) for all P
x=z iff P(x=y iff P(x) iff for all P) iff P(z)
x=a iff P(x=y iff P(x=y iff P(x=y iff P(x) iif P(y) for all P) iif P(x=y iff P(x) iif P(y) for all P) for all P) iif P(x=y iff P(x=y iff P(x) iif P(y) for all P) iif P(x=y iff P(x) iif P(y) for all P) for all P) for all P)
12:17
the small theorem of Goldbach :
Let n be a natural, non-zero integer, there exists a finite sequence of natural distinct integers, prime or equal to 1, the sum of which is n.
12:33
@Dattier wow, never heard of it :)
this is one of my enigmas
if the result existed before, I did not know it
@LeakyNun
12:50
what's a union of disjoint intervals?
a set?
it means you express the solution as something like [0,1] union (2,3]
hm.
oddly the hand out he gave was very very basic
and the excersizes he gave were very very hard.

« first day (2594 days earlier)      last day (2724 days later) »