DevelopmentLog.md

Development Log:

• Feb 9th (Hong-Ye Hu):

1. Add stabilizer.postselection() method
2. Add MeasureLayer() class. This will support mid-circuit measurement
3. Add Circuit() class. If there is no MeasurementLayer, self.unitary=True, and it should be the same as the CliffordCircuit class. If measurement is added, it is not unitary. forward(obj) will perform circuit unitary $U$ and measurements, and backward() will perform $U^{\dagger}$ and postselection on the states.
4. Add pyclifford quantum chemistry interface. The code is in qchem.py file. With PySCF and Qiskit, user can generate quantum chemistry Hamiltonian in pyclifford format, and calculated mean-field (Hatree-Fock) energy and exact energy.

• Feb 17th (Hong-Ye Hu):

1. Add many Pauli/Clifford gates, such as X,Y,Z,H,S,CNOT to circuit.py
2. Add 24 single qubit Clifford gate. And user can use C(gate_num, qubits) to specify the single qubit Clifford gate.

To Do List:

1. Check Circuit() is correct and replace CliffordCircuit.
2. Currently there is no conflict detection for CliffordGate class. So users can assign conflicting Clifford map by using both gate.set_generator() and gate.set_forward_map(). We need to add a conflict detection. (YZYou: there is a priority that when the gate implements the unitary transformation, generator will be used first, otherwise, clifford map.)
3. Speed up expectation() method: possible solution: 1) njit(parallel) 2) use Heisenberg evolution, and calculate expetation of Pauli strings in the zero state 3) test performance of Nvidia cuNumerics, to change numpy arrays.