I’m trying to understand and implement the amplitude amplification algorithm described in the HHL paper. I’m using the cirq implementation of the HHL algorithm as my starting point.
I have a couple of doubts with regard to this:
- The paper says the number of times the amplitude amplification procedure has to be applied to the HHL algorithm is and .
I’m unsure about how to calculate p˜ with regard to the cirq implementation. I’m also unclear as to how to find the functions f and g and the parameter λ˜. Could someone please help me understand this?
- With regard to the exact procedure to be used for amplitude amplification, the paper defines two new operators and
From what I understand, I need to loop over the HHL subroutine constantly applying an operator P = - RsuccRinit to it.
Could someone please explain how Rsucc and Rinit are described with respect to the cirq code? Can the P operator be expressed as another oracle?