# Curious Mathematica Result, Evaluating Bell Ineqality

1. Apr 28, 2010

### daverace

Hi guys, first post, sorry if it's in the wrong place. Frequent visitor, you seem to help pretty comprehensively.

I'm using mathematica to calculate a bell inequality, with a general noisy quantum state. However, when I evaluate it, I get a sequence with an a in the result.. which isn't a declared variable. I'm not really sure what it means, and given the generality of 'a', googling it is pretty much impossible. The question I'm asking is, what on earth is this 'a'? If it was 'i', I'd know how to start looking for a solution, but this is just strange.

I should be getting a value of 3.414 or so (maximum violation for a quantum state)

Here's an example of what I'm getting
Code (Text):
Evaluate[B1,q=1]
Sequence[2.28222 + 0.0266136 a, 1]

2. Relevant equations

Here's the equations I'm using, if they're at all relevant.
Code (Text):

x = ArcTan[q]
A = (Cos[x/2]^2)/2
B = (Sin[x/2]^2)/2
ab00 = A (2 - q)
ab01 = B (2 - q)
c = (Cos[x/2] + Sin[x/2])/4
d = (Cos[x/2] - Sin[x/2])/4
B1 = ab00 + ab00 + 2 c (q) + 2 q (A) + 2 a (B) + 2 q (d)

3. The attempt at a solution

No attempted solution, just too confused!

2. Apr 28, 2010

### kuruman

I can't claim to understand exactly what you are doing here, but in the last line should you have two ab00 terms instead of one ab00 and one ab01? If yes, why bother defining ab01?

3. Apr 28, 2010

### daverace

The actual terms for the X0Y0 measurements and X0Y1's are the same (hence me using ab00 twice). (edit:) The ab01 is used elsewhere in the calculations, didn't realised I'd copied it twice, by accident.

It's not the exact terms that's the problem, repetition, whether I've got the Bell Inequality right etc.

It's why has this 2.0.. + 1.00a, that a, turned up in the answer?

4. Apr 28, 2010

### daverace

Haha, nevermind. I typoed the a. Man I'm stupid for not checking that. Sorry for wasting your time there :]

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook