Main Page | Report this Page
 
   
Science Forum Index  »  Logic Forum  »  a Simple definition for 'finite'
Page 1 of 2    Goto page 1, 2  Next
Author Message
Guest
Posted: Thu Apr 24, 2008 12:05 pm
Hi all,

Define:
x is finite <-> ER Ay (y non empty subset of x ->
Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))).

were R is a strict total ordering on x.
i.e R is asymmetric,transitive and total binary relation on x.

In words: x is said to be finite if and only if there exist a strict
total order R on x such that every non empty subset y of x is
R_bi-ended.

y is R_bi-ended iff Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

So the above definition will be reduced to

x is finite iff ER Ay (y non empty subset of x -> y is R_bi-ended).

***********
e : epsilon membership
E: existential quatifier
A: universal quantifier


Zuhair
Guest
Posted: Thu Apr 24, 2008 3:56 pm
On Apr 24, 3:05 pm, Zaljo...@gmail.com wrote:
Quote:
Hi all,

Define:
x is finite <-> ER Ay (y non empty subset of x -
Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))).

were R is a strict total ordering on x.
i.e R is asymmetric,transitive and total binary relation on x.

In words: x is said to be finite if and only if there exist a strict
total order R on x such that every non empty subset y of x is
 R_bi-ended.

y is R_bi-ended iff Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

So the above definition will be reduced to

x is finite iff ER Ay (y non empty subset of x -> y is R_bi-ended).

***********
e : epsilon membership
E: existential quatifier
A: universal quantifier

Zuhair

What if we stipulate that R is a total relation on x.

R is total on x <-> Amn((mex nex) -> mRn or nRm)

y is R_bi-ended <-> Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Define: x is finite <->
ER Ay (y non empty subset of x -> (y is R_bi-ended & R is total on
y)).

would that definition be equivalent to the standard definition of x is
finite , I mean Tarski's definition of equinumerousity to a natural
number?

Zuhair
Guest
Posted: Fri Apr 25, 2008 1:17 am
On Apr 25, 1:12 am, William Elliot <ma...@hevanet.remove.com> wrote:
Quote:
On Thu, 24 Apr 2008 Zaljo...@gmail.com wrote:

In words: x is said to be finite if and only if there exist a strict
total order R on x such that every non empty subset y of x is
 R_bi-ended.

y is R_bi-ended iff Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Huh?  Let me guess, rather thatdecipherthewriter'scramp.

Define: y is R_bi-ended iff exists m exists n ( m in y and
n in y and notexist c ( c in y and c R m ) and
notexist c ( c in y and n R c ) )

Define: R is total on y iff
forall m forall n ((m in y and n in y) implies (m R n or n R
m))


Define: x is finite iff
exists R forall y (y non empty subset of x implies
(y is R_bi-ended and R is total on y)).

Is that clear to you now?

Zuhair


Quote:

A set S is finite when there is a linear order for S
with for all nonnul A subset S, min A and max A in A.

A set S is finite when there's a linear order
for S with for all A subset S, A is compact.

e : epsilon membership
E: existential quatifier
A: universal quantifier

's e t' is 's in t' while 'set' is set, like set up.

Riddle of the day.  Do you have to be a member to get into a set?

----
William Elliot
Posted: Fri Apr 25, 2008 3:12 am
Guest
On Thu, 24 Apr 2008 Zaljohar@gmail.com wrote:

Quote:
In words: x is said to be finite if and only if there exist a strict
total order R on x such that every non empty subset y of x is
R_bi-ended.

y is R_bi-ended iff Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Huh? Let me guess, rather thatdecipherthewriter'scramp.


A set S is finite when there is a linear order for S
with for all nonnul A subset S, min A and max A in A.

A set S is finite when there's a linear order
for S with for all A subset S, A is compact.

Quote:
e : epsilon membership
E: existential quatifier
A: universal quantifier

's e t' is 's in t' while 'set' is set, like set up.


Riddle of the day. Do you have to be a member to get into a set?

----
William Elliot
Posted: Fri Apr 25, 2008 3:57 am
Guest
On Fri, 25 Apr 2008, William Elliot wrote:

Quote:
On Thu, 24 Apr 2008 Zaljohar@gmail.com wrote:

In words: x is said to be finite if and only if there exist a strict
total order R on x such that every non empty subset y of x is
R_bi-ended.

y is R_bi-ended iff Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Huh? Let me guess, rather thatdecipherthewriter'scramp.

A set S is finite when there is a linear order for S
with for all nonnul A subset S, min A and max A in A.

A set S is finite when there's a linear order
for S with for all A subset S, A is compact.

A set is finite iff it can be given a Hausdorff

topology with the property
.. . for all A subset S, A is compact.

A set is finite iff it's a compact metric space
with the discrete metric.

Quote:
e : epsilon membership
E: existential quatifier
A: universal quantifier

's e t' is 's in t' while 'set' is set, like set up.

Riddle of the day. Do you have to be a member to get into a set?

----
MoeBlee
Posted: Fri Apr 25, 2008 8:48 am
Guest
On Apr 24, 6:56 pm, Zaljo...@gmail.com wrote:

Quote:
R is total on x <-> Amn((mex nex) -> mRn or nRm)

y is R_bi-ended <-> Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Define: x is finite <-
ER Ay (y non empty subset of x -> (y is R_bi-ended & R is total on
y)).

would that definition be equivalent to the standard definition of x is
finite , I mean Tarski's definition of equinumerousity to a natural
number?

It seems to me your definition entails that {0} is not finite.

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

MoeBlee
William Elliot
Posted: Sat Apr 26, 2008 12:20 am
Guest
On Fri, 25 Apr 2008 Zaljohar@gmail.com wrote:
Quote:
On Apr 25, 1:12 am, William Elliot <ma...@hevanet.remove.com> wrote:
On Thu, 24 Apr 2008 Zaljo...@gmail.com wrote:

Define: y is R_bi-ended iff exists m exists n ( m in y and
n in y and notexist c ( c in y and c R m ) and
notexist c ( c in y and n R c ) )

A subset (S,<=) is bi-ended when some a,b in A
with for all x in A, a <= x <= b. For example [0,1].

Quote:
Define: R is total on y iff
forall m forall n ((m in y and n in y) implies (m R n or n R
m))

Interesting notion, a relation with just tricotomy.

Tricotomy however isn't needed.

Quote:
Define: x is finite iff
exists R forall y (y non empty subset of x implies
(y is R_bi-ended and R is total on y)).

A set S is finite iff some (partial) order <= for S with
. . for all nonnul A subset S, A is bi-ended.

Quote:
Is that clear to you now?

What's R? Just an any old thing?
Maybe, if you made it clear that you expect telepathy
of your readers, R is simply a tricotomic relation?

Exercise. If <= is an asymmetric transitive relation for a set S
with for all nonnul A subset S, A is bi-ended, then <= is linear order.

Quote:
Riddle of the day. @Do you have to be a member to get into a set?

Riddle of the day. Is set therory set in its ways?
William Elliot
Posted: Sat Apr 26, 2008 12:39 am
Guest
On Fri, 25 Apr 2008, MoeBlee wrote:
Quote:
On Apr 24, 6:56 pm, Zaljo...@gmail.com wrote:

<snip, unreadable eye strain>

Quote:
It seems to me your definition entails that {0} is not finite.

It does? How do you read his cramedtogetherstuff?


See my answer to OP where I give a definition using orders
for finite that makes {a} finite.

Quote:
Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

Likely not as he has often multiple posts correcting his post even before
he gets a reply. Now on the other hand, my new and improved finite which
knows singletons are finite, swears up and down that empty set is
infinite.

That is useful for proving inf empty = oo and sup empty = -oo. ;-)

----
Guest
Posted: Sat Apr 26, 2008 4:37 am
On Apr 25, 11:48 am, MoeBlee <jazzm...@hotmail.com> wrote:
Quote:
On Apr 24, 6:56 pm, Zaljo...@gmail.com wrote:

R is total on x <-> Amn((mex nex) -> mRn or nRm)

y is R_bi-ended <-> Emn (mey ney ~Ec(cey cRm) ~Ec(cey nRc))

Define: x is finite <-
ER Ay (y non empty subset of x -> (y is R_bi-ended & R is total on
y)).

would that definition be equivalent to the standard definition of x is
finite , I mean Tarski's definition of equinumerousity to a natural
number?

It seems to me your definition entails that {0} is not finite.

hmmm...., yes you are right!

R is z-total <-> Amn((mex nex ~m=n) -> mRn or nRm)

Define:

Define: x is finite <->
ER Ay (y non empty subset of x->(y is R_bi-ended & R is z-total on y))

Now take the relation < as defined on natural numbers
then {0} is R_bi-ended.

Zuhair

Quote:

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

yes few seconds yes.


> MoeBlee
William Elliot
Posted: Sat Apr 26, 2008 5:48 am
Guest
On Fri, 25 Apr 2008, William Elliot wrote:
Quote:
On Fri, 25 Apr 2008, MoeBlee wrote:
On Apr 24, 6:56 pm, Zaljo...@gmail.com wrote:

It seems to me your definition entails that {0} is not finite.

See my answer to OP where I give a definition using orders
for finite that makes {a} finite.

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

Likely not as he has often multiple posts correcting his post even
before he gets a reply. Now on the other hand, my new and improved
finite which knows singletons are finite, swears up and down that empty
set is infinite.

That is useful for proving inf empty = oo and sup empty = -oo. ;-)

Whoops, another way will have to be found for proving that as

the new and improved finite, actually does include empty set.
MoeBlee
Posted: Mon Apr 28, 2008 8:26 am
Guest
On Apr 25, 10:39 pm, William Elliot <ma...@hevanet.remove.com> wrote:
Quote:
On Fri, 25 Apr 2008, MoeBlee wrote:

It seems to me your definition entails that {0} is not finite.

It does?  

Not just {0}, but any singleton.

Here's how (using Zuhair's definitions):

Definition:
R is total on x <-> AmexAnex(<m n>eR v <n m>eR)

Definition:
y is R_bi-ended <-> EmeyEney(~Ecey <c m>eR & ~Ecey <n c>eR)

Definition:
x is finite <-> ERAy(y is a non-empty subset of x -> (y is R_bi-ended
& R is total on y))

Suppose {z} is finite.
So, let Ay(y is a non-empty subset of {z}-> (y is R_bi-ended & R is
total on y)).
So {z} is R_bi-ended and R is total on {z}.
So let me{z} and ne{z} such that ~Ece{z} <c m>eR & ~Ece{z} <n c>eR.
So m=z=n. So ~<z z>eR.
But, since R is total on {z}, we have <z z>eR, which is a
contradiction.
So {z} is not finite.

MoeBlee



.








Quote:

See my answer to OP where I give a definition using orders
for finite that makes {a} finite.

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

Likely not as he has often multiple posts correcting his post even before
he gets a reply.  Now on the other hand, my new and improved finite which
knows singletons are finite, swears up and down that empty set is
infinite.

That is useful for proving inf empty = oo and sup empty = -oo. ;-)

----
Guest
Posted: Tue Apr 29, 2008 12:30 pm
On Apr 28, 11:26 am, MoeBlee <jazzm...@hotmail.com> wrote:
Quote:
On Apr 25, 10:39 pm, William Elliot <ma...@hevanet.remove.com> wrote:

On Fri, 25 Apr 2008, MoeBlee wrote:
It seems to me your definition entails that {0} is not finite.

It does?  

Not just {0}, but any singleton.

Here's how (using Zuhair's definitions):

Definition:
R is total on x <-> AmexAnex(<m n>eR  v <n m>eR)

Definition:
y is R_bi-ended <-> EmeyEney(~Ecey <c m>eR & ~Ecey <n c>eR)

Definition:
x is finite <-> ERAy(y is a non-empty subset of x -> (y is R_bi-ended
& R is total on y))

Suppose {z} is finite.
So, let Ay(y is a non-empty subset of {z}-> (y is R_bi-ended & R is
total on y)).
So {z} is R_bi-ended and R is total on {z}.
So let me{z} and ne{z} such that ~Ece{z} <c m>eR & ~Ece{z} <n c>eR.
So m=z=n. So ~<z z>eR.
But, since R is total on {z}, we have <z z>eR, which is a
contradiction.


I already corrected that. So what you say is not applicable to my
correction.
According to my modification on the definition of "total', it doesn't
follow that
if R is total on {z} then we have <z z>eR.

Zuhair

Quote:
So {z} is not finite.

MoeBlee

.





See my answer to OP where I give a definition using orders
for finite that makes {a} finite.

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

Likely not as he has often multiple posts correcting his post even before
he gets a reply.  Now on the other hand, my new and improved finite which
knows singletons are finite, swears up and down that empty set is
infinite.

That is useful for proving inf empty = oo and sup empty = -oo. ;-)

----- Hide quoted text -

- Show quoted text -
Guest
Posted: Tue Apr 29, 2008 1:12 pm
On Apr 28, 11:26 am, MoeBlee <jazzm...@hotmail.com> wrote:
Quote:
On Apr 25, 10:39 pm, William Elliot <ma...@hevanet.remove.com> wrote:

On Fri, 25 Apr 2008, MoeBlee wrote:
It seems to me your definition entails that {0} is not finite.

It does?  

Not just {0}, but any singleton.

Here's how (using Zuhair's definitions):

Definition:
R is total on x <-> AmexAnex(<m n>eR  v <n m>eR)

Definition

R is z-total on x <-> AmexAnex(~m=n -> <m n>eR v <n m>eR)

Quote:

Definition:
y is R_bi-ended <-> EmeyEney(~Ecey <c m>eR & ~Ecey <n c>eR)

Definition:
x is finite <-> ERAy(y is a non-empty subset of x -> (y is R_bi-ended
& R is total on y))

Suppose {z} is finite.
So, let Ay(y is a non-empty subset of {z}-> (y is R_bi-ended & R is
total on y)).
So {z} is R_bi-ended and R is total on {z}.
So let me{z} and ne{z} such that ~Ece{z} <c m>eR & ~Ece{z} <n c>eR.
So m=z=n. So ~<z z>eR.
But, since R is total on {z}, we have <z z>eR, which is a
contradiction.
So {z} is not finite.

MoeBlee

.





See my answer to OP where I give a definition using orders
for finite that makes {a} finite.

Do you play around with your newfangled proposals for at even just a
few seconds before you post them?

Likely not as he has often multiple posts correcting his post even before
he gets a reply.  Now on the other hand, my new and improved finite which
knows singletons are finite, swears up and down that empty set is
infinite.

That is useful for proving inf empty = oo and sup empty = -oo. ;-)

----- Hide quoted text -

- Show quoted text -
MoeBlee
Posted: Tue Apr 29, 2008 1:21 pm
Guest
On Apr 29, 3:30 pm, Zaljo...@gmail.com wrote:

Quote:
I already corrected that. So what you say is not applicable to my
correction.
According to my modification on the definition of "total', it doesn't
follow that
if R is total on {z} then we have <z z>eR.

Fine. And my post in response to the other poster was about my remark
that was based on your posts AT THAT TIME. Any subsequent changes you
made to your formulations have no bearing on what I proved based on
your formulations AT THAT TIME.

MoeBlee
Guest
Posted: Tue Apr 29, 2008 1:47 pm
On Apr 29, 4:21 pm, MoeBlee <jazzm...@hotmail.com> wrote:
Quote:
On Apr 29, 3:30 pm, Zaljo...@gmail.com wrote:

I already corrected that. So what you say is not applicable to my
correction.
According to my modification on the definition of "total', it doesn't
follow that
if R is total on {z} then we have <z z>eR.

Fine. And my post in response to the other poster was about my remark
that was based on your posts AT THAT TIME. Any subsequent changes you
made to your formulations have no bearing on what I proved based on
your formulations AT THAT TIME.

MoeBlee

Yes of course. it is a good point the one you've raised. and and it is
a good thing that you've writtin the proof. But I realized what you
said even without knowing the exact formal proof. It is clear that the
culprit was in the definition of 'total' that's why I changed it long
before you posted this proof.

Zuhair
 
Page 1 of 2    Goto page 1, 2  Next   All times are GMT - 5 Hours
The time now is Thu Jul 24, 2008 5:06 am