A silly computer experiment ... or, the heart of the matter?

Foundations of physics and/or philosophy of physics, and in particular, posts on unresolved or controversial issues

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 9:53 am

minkwe wrote:Do you believe it is possible to produce a 4xN spreadsheet of any size, from any source, introducing as much error as one likes, which produces a CHSH value above 2 ("statistically") by even 0.000000000001.

Jesus Christ. The answer is no. Understood by all but Joy Christian.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 9:58 am

Heinera wrote:
minkwe wrote:Do you believe it is possible to produce a 4xN spreadsheet of any size, from any source, introducing as much error as one likes, which produces a CHSH value above 2 ("statistically") by even 0.000000000001.

Jesus Christ. The answer is no.

You are aware that Richard argues that it can be violated by statistical error don't you. Richard argues that the CHSH is just a theoretical bound which can be violated a little due to statistical/experimental error. :shock: I'm happy you now see the issue. Nothing can violate the CHSH, not even non-locality or QM.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 10:12 am

minkwe wrote:
Heinera wrote:
minkwe wrote:Do you believe it is possible to produce a 4xN spreadsheet of any size, from any source, introducing as much error as one likes, which produces a CHSH value above 2 ("statistically") by even 0.000000000001.

Jesus Christ. The answer is no.

You are aware that Richard argues that it can be violated by statistical error don't you. Richard argues that the CHSH is just a theoretical bound which can be violated a little due to statistical/experimental error. :shock: I'm happy you now see the issue. Nothing can violate the CHSH, not even non-locality or QM.

No, he doesn't. A 4xN spreadsheet means that all correlations are measured on the same set of vectors. Richard is fully aware of the fact that the CHSH inequality is absolute in this case (he even gives a short proof in his paper. You should read it.).

What he says, is that if each correlation is computed on four different subsets, then the CHSH inequality can be violated. But then there is no 4xN spreadsheet.

He furthermore says: If the four different subsets are selected by a random process, then the violation can only be due to random errors. By increasing the size of the sets, we can get these random errors as close to zero as we want.
Last edited by Heinera on Sat Apr 12, 2014 10:15 am, edited 1 time in total.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 10:15 am

Heinera wrote:What he says, is that if each correlation is computed on four different subsets, then the CHSH inequality can be violated. But then there is no 4xN spreadsheet.

Funny, so you agree that Richard has been arguing that the upper bound of the CHSH inequality derived for a single set, applies to averages from 4 different sets, but only statistically?
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 10:20 am

minkwe wrote:
Heinera wrote:What he says, is that if each correlation is computed on four different subsets, then the CHSH inequality can be violated. But then there is no 4xN spreadsheet.

Funny, so you agree that Richard has been arguing that the upper bound of the CHSH inequality derived for a single set, applies to averages from 4 different sets, but only statistically?

The upper bound of the CHSH inequality derived for a single set, also applies to correlations from 4 different sets selected randomly, but only statistically.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 10:31 am

Heinera wrote:
minkwe wrote:
Heinera wrote:What he says, is that if each correlation is computed on four different subsets, then the CHSH inequality can be violated. But then there is no 4xN spreadsheet.

Funny, so you agree that Richard has been arguing that the upper bound of the CHSH inequality derived for a single set, applies to averages from 4 different sets, but only statistically?

The upper bound of the CHSH inequality derived for a single set, also applies to correlations from 4 different sets selected randomly, but only statistically.


Not only is this false, it demostrates that you too do not understand what upper bound means. Richard has since stopped talking about upper bounds because he realizes the silliness of suggesting that an upper bound is violated. Didn't you just confirm to me that the CHSH upper bound is never violated, even statistically, when applied to the system it was derived for. Upper bounds are mathematical tautologies, they do not care where you get your data. They can never be violated by anything.

If you want to prove that averages from different sets have a different upper bound, do that. Start from different sets and derive the upper bound from that. Why start with a single set to begin with like the CHSH. It is a mathematical crime. It is sad that you do not see it.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 10:40 am

You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 10:49 am

Heinera wrote:You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory.

You need to read up on probability. I just asked you if it was possible for the CHSH to be violated by a 4xN spreadsheet from any source, including as much error as you like, even by 0.0000001 and you said no. Now you are telling me that there is such a thing as an upper bound for expectations which can be violated?

Let us see: We have an expression:

E(U) + E(V) + E(X) + E(Y)

Those are expectation values, each has a maximum value of 1. Please tell me what you think the upper bound of that expression is. Can it be violated by experimental error.

Now we have another expression:

E(X) + E(-X) + E(Y) + E(-Y)

Please tell me what you think the upper bound for this expression is. Can it be violated by anything whatsoever?
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 11:02 am

minkwe wrote:
E(X) + E(-X) + E(Y) + E(-Y)

Please tell me what you think the upper bound for this expression is. Can it be violated by anything whatsoever?

No. But that's not the point. The point is:

E(X) + E(-U) + E(Y) + E(-V),
and
E(X) = E(U),
E(Y) = E(V).

Now, can E(X) + E(-U) + E(Y) + E(-V) be violated by anything whatsoever?
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 11:43 am

Heinera wrote:No. But that's not the point.

You can't even bring yourself to fully answer the simple questions

What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?

First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/experimental error. Note those are expectation values.

Remember we addressing your claim that:
You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.


Once we establish that, then we can go on to discuss specific possibilities of what X, Y, U, V, may be.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 12:04 pm

minkwe wrote:
Heinera wrote:No. But that's not the point.

You can't even bring yourself to fully answer the simple questions


Ok. Guess we have to reach for the teaspoon here.

Given that we assume
E(X) = E(U),
E(Y) = E(V),

I claim that

E(X) + E(-X) + E(Y) + E(-Y) = E(X) + E(-U) + E(Y) + E(-V).

Do you disagree?
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 12:31 pm

minkwe wrote:What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?

First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/experimental error. Note those are expectation values.

Remember we addressing your claim that:
You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.


Once we establish that, then we can go on to discuss specific possibilities of what X, Y, U, V, may be.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 12:43 pm

minkwe wrote:
minkwe wrote:What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?

First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/experimental error. Note those are expectation values.

Remember we addressing your claim that:
You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.


Once we establish that, then we can go on to discuss specific possibilities of what X, Y, U, V, may be.

That was quite a mouthful. I asked a yes/no question. A yes/no answer would be more productive.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 2:01 pm

Heinera wrote:That was quite a mouthful. I asked a yes/no question. A yes/no answer would be more productive.

So you know how to ask questions but you don't know how to answer them? I ask you a simple question and you bob and weave. A simple answer of my questions which I asked you first would have been more productive. Answer my simple questions which I asked you first, then you can ask me yours.

In case you forgot, let me remind you again: You said
Heinera wrote:You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.


So I asked you:

minkwe wrote:What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?

First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/experimental error. Note those are expectation values.


Can you or can you not answer those simple questions? Once you've answered, them you can then ask me whatever you want to ask.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 2:22 pm

minkwe wrote:So I asked you:

minkwe wrote:What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?.

I told you that the first could be violated (since U,V,X,Y could be anything), the second not. Do you feel that you can now answer my question?
Last edited by FrediFizzx on Sat Apr 12, 2014 2:32 pm, edited 1 time in total.
Reason: Removed Jesus comment.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 2:39 pm

Heinera wrote:
minkwe wrote:So I asked you:

minkwe wrote:What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?.

I told you that the first could be violated (since U,V,X,Y could be anything), the second not. Do you feel that you can now answer my question?

You haven't answered the questions. Don't you see that you have to answer what the bounds are first before you say whether it is violated or not. Please read the question carefully again and have another attempt because you haven't answered it.

minkwe wrote:First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/statistical error.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

Re: A silly computer experiment ... or, the heart of the mat

Postby Heinera » Sat Apr 12, 2014 3:01 pm

This is going absolutely nowhere: no answer from you. But you have certainly raised the dust regarding Joy's experiment. In another thread Joy is now obviously unsure of whether all correlations can be computed on the same set of vectors, or whether he needs some other protocol to succeed. At the same time, he wants to publish the list of vectors on the Internet, so that anyone can reproduce the QM correlations. But how? Minkwe says that they have to be very careful. You can't compute two correlations with different settings (a,b) on the same set. After the first computation, that data set is contaminated. Spooky, indeed. Way more interesting discussion. We continue in that other thread.
Heinera
 
Posts: 917
Joined: Thu Feb 06, 2014 1:50 am

Re: A silly computer experiment ... or, the heart of the mat

Postby FrediFizzx » Sat Apr 12, 2014 3:33 pm

Heinera wrote:This is going absolutely nowhere: no answer from you.

I really do think you should answer Michel's questions first which you have NOT answered. If you can't answer them, then just say so. No problem. Thanks.
FrediFizzx
Independent Physics Researcher
 
Posts: 2905
Joined: Tue Mar 19, 2013 7:12 pm
Location: N. California, USA

Re: A silly computer experiment ... or, the heart of the mat

Postby gill1109 » Sat Apr 12, 2014 5:11 pm

FrediFizzx wrote:
Heinera wrote:This is going absolutely nowhere: no answer from you.

I really do think you should answer Michel's questions first which you have NOT answered. If you can't answer them, then just say so. No problem. Thanks.


Fred, the problem is "ask a silly question, get a silly answer".
Heinera doesn't answer Michel's question because it is ill-posed.
There is a second major problem with is: this has gone totally off-topic.

Take one N x 4 spreadsheet of numbers +/-1 called A, A', B, B'

Calculate four correlations each based on all N numbers and you'll always find CHSH <= 2

Choose a random number 1, 2, 3, or 4 for each row of the spreadsheet, and re-order the rows of the spreadsheet by the new column (with numbers 1, 2, 3, 4). Call the numbers or rows with each value N1, N2, N3, N4; so N1 + N2 + N3 + N4 = N.

By copy-paste make four smaller spreadsheets, with N1, N2, N3 and N4 rows respectively

On small spread-sheet 1 calculate E(A, B)
On small spread-sheet 2 calculate E(A', B)
On small spread-sheet 3 calculate E(A, B')
On small spread-sheet 4 calculate E(A', B')

Calculate CHSH

About half the time, you could find CHSH <= 2
About half the time, you could find CHSH >= 2

You might on occasion find CHSH = 4

If N was large, however, you'll only rarely find CHSH > 2.4

Does anyone disagree?

Has everyone now done my little R experiment in order to confirm the things I say here, to get them clear in their minds?

Please all make a major effor to get back on topic, boys.

Do I have to say that in large bold capital letters?
gill1109
Mathematical Statistician
 
Posts: 2812
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: A silly computer experiment ... or, the heart of the mat

Postby minkwe » Sat Apr 12, 2014 5:19 pm

Richard,
Maybe you want to try to answer the questions, they are very simple, and clearly well-posed. You know the answers. All I'm asking is that one of you simply state the answers. Here it is again within the full context:
Heinera said:
Heinera wrote:You need to realize that the two words "upper bound" can be used in two different ways: Absolute upper bound, and upper bound for expectations. Read up on probability theory. When we are talking about an upper bound for expectations, the bound will usually be violated half of the time.


So I asked:
minkwe wrote:The expectation values E(U), E(V), E(X), E(Y), each have a maximum value of 1.

What is the upper bound for
E(U) + E(V) + E(X) + E(Y)

and what is the upper bound for
E(X) + E(-X) + E(Y) + E(-Y)

Can anything violate them whatsoever, even by experimental error?

First tell me what the upper bounds are, then tell me whether they can ever be violated by anything, including experimental/statistical error. Note those are expectation values.

Can you answer these questions Richard, or will you bob and weave just like he did. Unlike your claims, the questions are clearly not silly. The go to the root of you and Heinera's claim that "When we are talking about an upper bound for expectations, the bound will usually be violated half of the time".

In case you did not realize, this is on-topic. It demonstrates convincingly that your experiment claiming to show statistical violation of an upper bound is baloney.
minkwe
 
Posts: 1441
Joined: Sat Feb 08, 2014 10:22 am

PreviousNext

Return to Sci.Physics.Foundations

Who is online

Users browsing this forum: ahrefs [Bot] and 82 guests

cron
CodeCogs - An Open Source Scientific Library