This class uses the fact that the s_prime configurations will often (around 30% of the time) be multiplied by a matrixElement weight of 0, making them irrelevant. The EfficientPOVMOperator class will not evaluate the neural network for these configurations but just assign them a value of 0.
Implemented a more efficient
POVMOperatorclass.This class uses the fact that the s_prime configurations will often (around 30% of the time) be multiplied by a matrixElement weight of 0, making them irrelevant. The
EfficientPOVMOperatorclass will not evaluate the neural network for these configurations but just assign them a value of 0.