Convert a Qobj into numpy array

[Caco Somoza]

Hi there! I just started learning Qutip and my python skills are still novice. I was wondering if there's any Qobj method to convert an operator into a simple numpy array or related. I'm using Spyder since the variable explorer lets me see matrices in a friendly way (studying density matrices in the console is a nightmare). If you guys know any other way to display matrices inline in a neat way I would be very pleased to hear.

Thanks!

[Paul Nation]

To get a Qobj to an array use: 

Q.full()

Paul

On Mar 20, 2014, at 9:11 PM, Caco Somoza <cacos...@gmail.com> wrote:

Hi there! I just started learning Qutip and my python skills are still novice. I was wondering if there's any Qobj method to convert an operator into a simple numpy array or related. I'm using Spyder since the variable explorer lets me see matrices in a friendly way (studying density matrices in the console is a nightmare). If you guys know any other way to display matrices inline in a neat way I would be very pleased to hear.

Thanks!

--
You received this message because you are subscribed to the Google Groups "qutip" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qutip+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Vishvendra Punia]

Other useful operations on Quantum Objects. http://qutip.org/docs/2.2.0/guide/guide-basics.html

Function	Command	Description
Conjugate	Q.conj()	Conjugate of quantum object.
Dagger (adjoint)	Q.dag()	Returns adjoint (dagger) of object.
Diagonal	Q.diag()	Returns the diagonal elements.
Eigenenergies	Q.eigenenergies()	Eigenenergies (values) of operator.
Eigenstates	Q.eigenstates()	Returns eigenvalues and eigenvectors.
Exponential	Q.expm()	Matrix exponential of operator.
Full	Q.full()	Returns full (not sparse) array of Q’s data.
Groundstate	Q.groundstate()	Eigenval & eigket of Qobj groundstate.
Matrix Element	Q.matrix_element(bra,ket)	Matrix element <bra|Q|ket>
Norm	Q.norm()	Returns L2 norm for states, trace norm for operators.
Partial Trace	Q.ptrace(sel)	Partial trace returning components selected using ‘sel’ parameter.
Sqrt	Q.sqrtm()	Matrix sqrt of operator.
Tidyup	Q.tidyup()	Removes small elements from Qobj.
Trace	Q.tr()	Returns trace of quantum object.
Transform	Q.transform(inpt)	A basis transformation defined by matrix or list of kets ‘inpt’ .
Transpose	Q.trans()	Transpose of quantum object.
Unit	Q.unit()	Returns normalized (unit) vector Q/Q.norm().

[Joe Bowen]

Hi there,

I want to take the cosine of a Qobj. My way of doing this was to convert it to an numpy array then use np.cos(). How would I then change it back to a Qobj? I would appreciate any help or alternative ways of doing this.

Joe

[nonher...@gmail.com]

You would have to take the resulting array and pass it to the Qobj class

Out = Qobj(array)

Of course if you have tensor structure and other things these would have to be passed back to the class as well. 

Paul

--
You received this message because you are subscribed to the Google Groups "QuTiP: Quantum Toolbox in Python" group.

To unsubscribe from this group and stop receiving emails from it, send an email to qutip+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<qobj_numpy.py>

[Joe Bowen]

Yes, so my tensor structure in the original Qobj is;

        dims = [[10, 10], [10, 10]], shape = [100, 100], type = oper, isherm = True

I'm not sure how to pass the appropriate tensor structure back.

        Q.permute(order)

What would I pass into 'order'?

Thanks


[nonher...@gmail.com]

You can pass that information to the Qobj class just as you typed it. Or just use

Qobj.dims

Qobj.shape

etc from the old Qobj. 

Another possibility is to extract the array from the object and then set the resulting array after calculations as the Qobj data

my_array = Qobj.full()

Do stuff here

Qobj.data = sp.csr_matrix(array)

Where sp stands for the scipy.sparse module. 

--

[Alex Pitchford]

You could perhaps try subclassing to create your own Qobj with addition methods, e.g.

class QobjTrig(Qobj):

  def cos(self):

    out = Qobj()

    out.data = sp.csr_matrix(np.cos(self.data))

    out.dims = [self.dims[1], self.dims[0]]

    return out

This probably doesn't offer much over Paul's method unless you will be wanting to use the function over again and perhaps do some other trig functions as well.