|
|
View Full Version : Probability waves
Vkothii 10-19-08, 05:55 PM WTF is a wave of probability?
This can't be, surely, the model we use to explain fundamental particles??
Are there any experts here who know what this rather peculiar phrase means?
Or is it just pseudoscientific mumbo-jumbo that the "experts" would have us believe actually does explain something?
Is probability real? How is it connected to numbers, since numbers are imaginary?
Translation: how do we assign or measure probabilities, and how do they "wave" or oscillate, and around what?
BenTheMan 10-19-08, 06:06 PM http://en.wikipedia.org/wiki/Wavefunction and references therein.
CheskiChips 10-19-08, 06:08 PM If there's a clearly oscillatory pattern for a particle, but the individual constituent parts can't be quantified you use probability. Thus it's notated as a wave with having the 'probability' of any given state given any set of circumstances at any given time.
Since you're interested in information-theory; it's interesting to note that the probabilities can be defined as packets of information with unknown causation at this time. The inconsistencies would be as unique as perhaps a full DNA strain...which is not very unique at all.
Vkothii 10-19-08, 06:09 PM If there are such things, are they real, these "probability waves", they exist and have real values?
Is an apple a real value, or a real object?
Ben: if you can't see it, what I'm doing here is working backwards from a conclusion to a premise. Maybe you can recall what the premise was, the one you made?
BenTheMan 10-19-08, 06:27 PM If there are such things, are they real, these "probability waves", they exist and have real values?
Is an apple a real value, or a real object?
Ben: if you can't see it, what I'm doing here is working backwards from a conclusion to a premise. Maybe you can recall what the premise was, the one you made?
I don't understand anything anymore.
I had assumed that you were doing more than cutting and pasting formulae from Wikipedia in your other thread, and now I see that that's all you were doing. So much for me giving people the benefit of the doubt.
The wavefunction has the interpretation of a probability wave, i.e. a probability distribution. In general it is a complex number. However, what you learn very quickly when you study quantum mechanics is that the wavefunction is not physical---in order to get a probability, you have to integrate the absolute square of the distribution over all space, in other words
P = \int \left|\psi\right|^2.
Now, the quantity \left|\psi\right|^2 is manifestly real. If you'd like a demonstration of this, I direct you to the following resource: http://en.wikipedia.org/wiki/Complex_numbers.
Vkothii 10-19-08, 07:12 PM the wavefunction is not physicalYou mean "we can't measure a wavefunction"?
But it is a 'real' thing, surely? How can it not be real?
the quantity \left|\psi\right|^2 is manifestly real.When all you have to do to make it 'real' is measure a complex amplitude?
Could there be a bit of an ontology snafu here?
How does something with no physical reality (a wavefunction) "produce" something that has a physical reality (a probability). Is it something to do with measurement?
I thought you might have some understanding of the measurement problem, but I see that you would rather stick to a particular method, which looks a lot like "if I just repeat myself, it will start to make sense eventually".
Vkothii 10-19-08, 10:07 PM How does one measure something? How do we measure the strength of a magnetic field?
We have to interact with the sucka. Use something that moves or blinks when we 'put' it in there.
If we have an independent measurement of local field strength, we know that things like electrons' spin will move or blink a certain way - this is our expectation.
When a beam of coherent light is plane-polarized, there's a wavefunction describing the geometric polarization as a probability of finding a photon polarized in the plane, or at a perpendicular plane (that is orthogonal) direction. There is also a wavefunction describing the geometry of the polarization as two oppositely circular-polarized photons.
Where is the information content? Has any been 'put' in there, by polarizing a beam of light?
Vkothii 10-20-08, 07:54 PM The answer is, yes of course there is more information in a polarized photon.
You just have to 'measure' the content; this requires a 'measurement basis'.
You can't analyze a single polarizing filter, as a kind of sieve, but you can if you have two filters (or more than two). Why would that be??
It's because the first polarizing filter 'computes' a plane-polarized 'wavefunction', which additional filters can 'measure' (by computing the existing polarity of each photon [from the first filter]). Thus, the additional filters can process (i.e. compute) an existing informational entropy, the one 'computed' by the first filter.
Although photons have to be analysed "non-classically" when they get processed by filters, beam-splitters and mirrors, as wavefunctions with probability amplitudes, they can be encoded; the 'information signal' is polarity (of the wavefunction), in this case.
BenTheMan 10-20-08, 08:28 PM When all you have to do to make it 'real' is measure a complex amplitude?
No. The thing you measure is always real.
Vkothii 10-20-08, 11:23 PM Bingo.
You asked: "is a probability a real number?"
The answer is: no, a real number is a number, that we assign to a real probability. As you say, we 'measure' things which are real and physical (have a real extent in space and time).
So strictly speaking a probability is a real possibility of an actual event, the real number corresponding to any possible event is always a number between zero and one.
Therefore information is both logical (as in real numbers), and physical.
Vkothii 10-21-08, 05:03 AM Part (ii), in electrons, nucleons and massless photons, this probability of being spin up or down, or polarity of spin-[1] photons, which they exhibit as polarization wavefunctions, along a dimensional axis, including parallel to the dimension of propagation (i.e. the positive z-axis), but fermions exhibit as superpositions of spin wavefunctions:
how do [probabilities] "wave" or oscillate, and around what?
It has something to do with Taylor's expansion, with imaginary exponents (by which I mean\, i rotates a number into the Euclidian plane of imaginary roots).
Something to do with algorithms too, and computation. Computations are eigenstates, and eigenvalues or measurements
But what do these probabilities do? Who can say?
Like I've been rabbiting away about information and communication (i.e. measurement of a signal from somewhere), it depends what you define as the sender/receiver, and which is the channel that commutes the signal - the alphabet. If the sender is an electron that 'spins' into a new state wavefunction-wise (i.e wrt to a nucleus) and 'sends' a photon message, then that's the basis of the measure-space, is one way to say "how to send the signal'. If it's photons as the quantum, then they 'send' matter-waves as electrons, or they encode electrons.
Because we decide, fundamentally, which is which when it comes to the content.
BenTheMan 10-21-08, 12:26 PM The answer is: no, a real number is a number, that we assign to a real probability.
What? What about the real number 2?
Vkothii 10-21-08, 02:46 PM What about it?
2 is a number, right? So is 2.141267, right?
If I have 2 apples, what do I have? A number, or something I can eat?
(This is getting a bit strange, once again - you do understand that numbers are things we count with, etc, you can't "do anything" with a number. Except attach it to something - an abstraction. You understand what abstraction is, or didn't they tell you about that one??)
AlphaNumeric 10-21-08, 03:00 PM I f I have 2 apples, what do I have? A number, or something I can eat?You have something which you can eat which also has a property which represents the concept of 'Two-ness', which is a property shared by all other pairs of objects/entities
by which I mean 'i' rotates a number into the Euclidian plane of imaginary rootsNo, given a number z in the Complex plane (which is not a metric space so the term 'Euclidean' should not be used), iz can be obtained by rotating z about the origin by \frac{\pi}{2} radians.
Please, please, please can you refrain from trying to put your own 'spin' on things you aren't sure about. If you don't know, don't try to fill in the gaps. Not for myself but so you avoid confusing people who don't know about complex numbers but whom are reading this thread. That's part of the reason I don't let mistakes slide so readily, if this forum exists to help educate some or all of its readers and posters, trying to fill in the gaps is worse than just saying nothing about a topic.
You asked: "is a probability a real number?"
The answer is: no, a real number is a number, that we assign to a real probability. As you say, we 'measure' things which are real and physical (have a real extent in space and time).Ignoring metaphysics for a moment, probability is a Real (I use capital R to show it's not being used in the same way as Pinocchio says "I'm a real boy" but to refer to the field of real numbers (http://en.wikipedia.org/wiki/Real_number). A probability is a Real number between, and including, 0 and 1. Not all Reals can be probabilities but all probabilities are Reals.
So strictly speaking a probability is a real possibility of an actual event, Circular definition. You cannot use a word to define itself.
Vkothii 10-21-08, 03:03 PM A probability is not a number. It's a real probability.
Probability is 'numerical' but it's real, not an abstraction. Real things are probabilities, not abstractions.
A plane is Euclidean.
Please, please, don't have a tantrum.
Vkothii, these funtions describe how particles will behave in a probabilistic way. They are "real" in the sense that particles really do behave in a wave-like manner, and these functions can accurately describe that behavior. Asking if the functions themselves are "real" is a lot like asking if force=mass x acceleration is "real". Perhaps the best way to put it would be to say "it accurately describes something that really happens".
Vkothii 10-21-08, 04:24 PM Perhaps the best way to put it would be to say "it accurately describes something that really happens". Really?
So a number isn't a probability, then? Real probabilities instead are abstracted logically to real 'numbers', and real physical events and objects are abstracted to logical 'information', then?
There are all these 'formulas' made out of mathematics, type of thing?
Vkothii 10-21-08, 07:12 PM You could even get all carried away and think something like: "a point can be information, a number, because it can describe a position."
If you want to 'send' or 'transmit' this single bit of positional information, you absolutely require another dimension, so information is dimensional, and it isn't 'one-dimensional' or static, either.
To get a point from one position to another position, you translate it in another dimension, or 'project' the point, and make a line. This one-dimensional line is the integral of all positions the zero-dimensional point can be in, wrt the start and end positions, so each point can be assigned a position (a number) along this 1-dimensional object. You can call this object "one line", or a line that is "one unit" of information.
Actually then, unless a point has a position, it is 'informationally zero', or it's 'dimension' of information is zero.
As soon as you give it a number, it has a dimension which is a position in a 1-dimensional space.
This encapsulates the idea of having to project something (a physical position), to get it to be information (a number which is logical, not physical).
You give something an informational dimension, by translating, or projecting it, generally.
AlphaNumeric 10-22-08, 03:28 AM Probability is 'numerical' but it's real, not an abstraction. Real things are probabilities, not abstractions.So can you point at a '1 in 5' dog. Or a 2 in 7 table? I can point at 4 dogs or 2 apples. Why are probabilities, which are a measure of the ratio of numbers, more real than the numbers themselves?
A plane is Euclidean.
No, 'Euclidean' is a specific type of space which has a metric or inner product defined upon it. Given a 2 dimensional space, you cannot say if it's Euclidean or not (say Lorentzian) till I tell you how to measure distances between points on that plane. If I tell you that the distance, s, from (0,0) to (x,y) is measured as s^{2} = x^{2}-y^{2} it is clearly not Euclidean, despite the (x,y) plane being 'flat'. There's a great many other ways of measuring distances, which you would know if you knew about 'metric spaces'. Which you don't.
Is this going to be another thing you can't accept you're wrong on or are you going to actually admit you're incorrect on this one? Hopefully the fact I've given an example of a non-Euclidean plane will not fall on deaf ears...
Vkothii 10-22-08, 03:31 AM So can you point at a '1 in 5' dog. Or a 2 in 7 table? I can point at 4 dogs or 2 apples. Why are probabilities, which are a measure of the ratio of numbers, more real than the numbers themselves?Can you point at a dog and not identify it as "a dog"? Is there a probability you will call it "a horse"? What value is this probability, in numerical and in social terms? Does this probability have a value in psychological terms?
There's a great many other ways of measuring distances, which you would know if you knew about 'metric spaces'. Which you don't.I know what a "metric" is though.
I know that it can be whatever we say it is, for example, which you don't.
AlphaNumeric 10-22-08, 06:05 AM Can you point at a dog and not identify it as "a dog"? Is there a probability you will call it "a horse"? What value is this probability, in numerical and in social terms? Does this probability have a value in psychological terms?How does that make probability more actual than numbers?
I know what a "metric" is though. And yet you didn't realise that planes are not automatically Euclidean. The 2d plane doesn't have a metric automatically assigned to it.
know that it can be whatever we say it is, for example, which you don't. No, for instance a metric for \mathbb{R}^{2} cannot be \left( \begin{array}{cc} -1 & 1 \\ 0 & 1\end{array} \right) in ANY basis. Tell me why not.
Vkothii 10-22-08, 06:18 AM Do you mean "why not mathematically", or "why not logically"?
You realize I just have to invent a new logic, and apply it? This would not be cheating, being an entirely arbitrary, therefore logical encoding to use?
That's what "it can be whatever we say it is" means.
And, again, how are numbers more real than probabilities?
QuarkHead 10-22-08, 12:06 PM Do you mean "why not mathematically", or "why not logically"? If you truly believe there is a difference, why not do it both ways? Should be easy.......
Vkothii 10-22-08, 12:25 PM What is the probability that I can find >1 possible ways of representing that 2x2 matrix, or the "reals", in math-speak, to get a solution that is logically consistent, though?
Seeing how I'm trying to find out what probability is.
Correction, I'm trying to find out if anyone else understands what it is. So far, there's this theory that probability is a real number; or it's real, and we give it something called "a number".
This appears to hang on what someone thinks a number might actually be.
But the math-dudes want to play games with logic instead, or something.
I think I can say I'm happier with the idea that numbers are abstractions, probabilities are real events.
You can stick with your one about numbers being real, and probabilities are the abstraction, by all means.
Just like I can invent a mathematical representation that proves that the little exercise in math, is actually an exercise in logic,
Or didn't I do that already, by saying there's no reason I can't? So the "exercise" is trivial. And so is my belief that real events determine probable outcomes, numbers don't do anything, because numbers are abstractions.
But you don't have to believe this, or anything else either.
P.S. Here's a logic puzzle for yez: If any real physical thing is in fact, some kind of processor (a rock processes its environment, a drop of water processes any other chemicals that "dissolve" in it, etc), then is a number a process? Do numbers compute anything? What is the probability that two numbers can compute a third?
Is it approximately 1/3? Or 0?
Vkothii 10-23-08, 12:19 AM Yes, well.
Obviously, if you stretch the notion of what a number is, and assume a function that generates a value, is a number, then numbers are functional too.
If you accept that numbers are 'ergodic', rather than whatever it is that drives these real-valued functionals to give up a value, surrender it to our grasp.
Things like probabilities, too, must be functional since a probability is a prediction as well as an outcome.
Probability 'functions' like a circuit, with a steady-state response of 'prediction', and a perturbed condition of 'real result'.
But this is because computers run algorithms, after being given lots of other numbers.
Probabilities occur randomly as real events, but don't predict anything.
That's our job.
You just have to ensure you use it mathematically, and imply that the real event is predicted by a mathematical 'measure', a function that computes some value. But the real event still just 'happens', it doesn't need mathematics to do it.
Vkothii 10-23-08, 03:08 AM Let's see if I can get this one rejected by the "science censors".
The entropy of a partition is closely related to the idea of the minimum amount (of information) that can be known about some evolving system (a heat engine, say), in terms of a very small, or smallest portion of the overall 'interaction space'.
So we have Louiville's eqn. and a description of a distribution of a canonical position and momentum, over a phase space. The idea is to project this model into the space of interactions between fundamental particles like electrons and photons, while ensuring the volume of the space is conserved.
That means probability is fundamentally conserved.
"This is for those advanced thinkers who may (or may not) have thought about what a computer actually is."
You actually seem to think you do know. This must be something that bothers you, since you seem to be prepared to go to the trouble of actively turfing such troublesome ideas away.
Vkothii 10-23-08, 04:55 AM There really is only one way to see it.
You have to have a sender and a receiver, for a message to be possible, it doesn't get much simpler.
Real space and time has real things in it, and they move around. You can logically assign an information content to any object and/or movement, it's called a representation.
So everything, in the entire universe that we can see, is real and informational as well; in short, information is physical, light is a message from physical objects, just the same way a waveform traveling down a wire encoded with 'speech' is a message from a telephone.
There is no fundamental difference between saying the radiation we get every day is heat and light across the spectrum, and saying it's the information we get.
This is the guts of Information Theory, that not just electronic circuits are channels encoded with signals. Everything that can be identified as having an input, an output, and transforming something from one to the other, possibly both ways. Lots of things behave like a sending/receiving circuit - an information channel.
Apparently no-one else at this forum has heard of it, but then it is only more than 50 years old.
Thinking that the idea of a distant 'light' from a distant celestial object is communication, must be a dumb idea, is up there with thinking that communicating with a lighthouse or a Very lantern, or by sending light down glass fibers say, is a dumb idea.
bestofthebest 06-14-09, 04:17 PM probabilty waves are real to the extent needed to make any rational sense of the quantum world
strictly speaking they aren't waves as such but it's hard to describe what the really just like everything else in quantum phisics and a waves happens to be the most convienient way of looking at it because one we get to the different between bosons and fermions the whole difference is that when to identical fermions interact one of the probabilty waves is flipped meaning it is impossible for to fermions to be in exactly the same state at once (eg. two electrons with the same spin cannot occupy the same position in space)
probabilty are to an extent some mumbo-jumbo that is there to help us understand but if it wasn't there quantum phisics would be even harder
like richard feynman said "i think i can safely say nobody understands quantum phisics"
so don't look into it too much but i recommend reading "quantum theory cannot hurt you" by Marcus Chown, it explains alot.
|