(cont.) Ok, here's what I am trying to model:
I have a double slit apparatus (in fact, any apparatus with some observables that can give an interference pattern will do), which constantly firing electrons towards the screen, thus the computer will be recording some interference pattern building up.
To this apparatus, I also have the electron being entangled with a qubit located in a trap some meters away from the apparatus (so as not to put two bulky things too close)
Therefore I have the following entangled state, where stuff on the left side of the $\otimes$ corresponds to the qubit and the superposition on the right side of the $\otimes$ corresponds to the which slit observable in the double slit apparatus. i.e.
$\lvert s\rangle = \frac{1}{\sqrt{2}}(\lvert 0\rangle \otimes \frac{1}{\sqrt{2}}(\lvert 0\rangle + \lvert 1\rangle) + \lvert 1\rangle \otimes \frac{1}{\sqrt{2}}(\lvert 0\rangle - \lvert 1\rangle))$
$=\frac{1}{2}(\lvert 00\rangle + \lvert 01\rangle + \lvert 10\rangle - \lvert 11\rangle)$
when wrote as a density matrix we get:
$\lvert s\rangle\langle s\rvert = \frac{1}{4}\begin{pmatrix}1 & 1 & 1 & -1\\1 & 1 & 1 & -1\\1 & 1 & 1 & -1\\-1 & -1 & -1 & 1\end{pmatrix}$
$\lvert s\rangle = \frac{1}{\sqrt{2}}(\lvert 0\rangle \otimes \frac{1}{\sqrt{2}}(\lvert L\rangle + \lvert R\rangle) + \lvert 1\rangle \otimes \frac{1}{\sqrt{2}}(\lvert L\rangle - \lvert R\rangle))$
$=\frac{1}{2}(\lvert 0L\rangle + \lvert 0R\rangle + \lvert 1L\rangle - \lvert 1R\rangle)$
(and the density matrix is the same as it is written under the basis generated by $\{\lvert 0L\rangle, \lvert 0R\rangle, \lvert 1L \rangle,\lvert 1R\rangle\}$
We can easily take a partial trace wrt the qubit to confirm that we will get 50:50 on which slit the electron passes through
$\frac{1}{4}\begin{pmatrix} 2 & 0\\ 0 & 2\end{pmatrix} = \begin{pmatrix} \frac{1}{2} & 0\\ 0 & \frac{1}{2}\end{pmatrix}$
Now I am preparing to measure the spin of the qubit along the z direction. The measuring device, being macroscopic, has many degrees of freedom, hence it is basically like an environment to the system in question
Just before measurement, I should have the following system-environment state:
$\lvert\varepsilon \rangle\otimes \lvert s\rangle = \frac{1}{\sqrt{2}}\lvert\varepsilon\rangle \otimes ( \lvert 0\rangle \otimes \frac{1}{\sqrt{2}}(\lvert L\rangle + \lvert R\rangle) + \lvert 1\rangle \otimes \frac{1}{\sqrt{2}}(\lvert L\rangle - \lvert R\rangle))$
$=\frac{1}{2}(\lvert \varepsilon 0L\rangle + \lvert \varepsilon 0R\rangle + \lvert \varepsilon 1L\rangle - \lvert \varepsilon 1R\rangle)$
Then the environment absorbs the system's states and give:
$=\frac{1}{2}(\lvert \varepsilon_{0}L\rangle + \lvert \varepsilon_{0}R\rangle + \lvert \varepsilon_{1}L\rangle - \lvert \varepsilon_{1}R\rangle)$
Thus by einselection and decoherence, $\langle \varepsilon_i\vert \varepsilon_j\rangle = \delta_{ij}$
However, what I am not sure now, is whether I have the correct system-environment state to query the statistics of the double slit experiment after the measurement of the qubit is being done. By unitarity and no communication theorem, a local operation at one subsystem of the entangled system will not change the reduced density matrix of the other subsystem, so I should still have the interference pattern building up unaltered before and after measurement?