Noise transform

[gadial]

Summary

This module transforms one error channel into another built from specified building blocks, such that the two channels are as close as possible. e.g. converting a general noise to a Clifford noise that a Clifford simulator can handle.

Details and comments

This module adds two functions:

• approximate_quantum_error: given an input channel and a description of the desired building blocks, and output a "close" channel built from the building blocks.
• approximate_noise_model: applies approximate_quantum_error to all noises in the noise model, with the same set of building blocks used for all noises in the model.

Examples:

Inputs as quantum error and operator string:

gamma = 0.23
error = amplitude_damping_error(gamma)
results = approximate_quantum_error(error, operator_string="reset")

Inputs as Kraus matrices and operator list:

X = standard_gate_unitary('x')
Y = standard_gate_unitary('y')
Z = standard_gate_unitary('z')
p = 0.34
theta = numpy.pi / 7
E0 = numpy.sqrt(1 - p) * numpy.array(numpy.eye(2))
E1 = numpy.sqrt(p) * (numpy.cos(theta) * X + numpy.sin(theta) * Y)
results = approximate_quantum_error((E0, E1), operator_list=[X, Y, Z])

[chriseclectic]

Looks like something odd is going on with the math operator font characters here

[chriseclectic]

This looks good @gadial, thanks!

The only minor issue I see now is that the tensor product and compose math operators have been replaced with question marks in the quantum_error.py file. If you can fix that its ready to merge (and if it's too hard to fix because of a font issue on your computer I can fix it in a seperate PR)

[gadial]

Should be ok now.

[chriseclectic]

Great work @gadial !