optimizer.jl is a small Julia library for finding optimized n to k purification circuits. It makes use of a Monte Carlo simulator, which allows for a large speed up. Note that the calculated fidelities are not exact, and their error scales with
examples/run_experiment.jl contains an example run through of using the optimizer. An example output from an experiment run is located in examples/data/example_output.csv. examples/plotting_example runs through how some of the data can be visualized. examples/submit.sh is an example of what can be used for running simulations with a job array and multithreading on a supercomputing cluster.
run_experiment.jl contains an example run through of using the optimizer. submit.sh is an example of what can be used for running simulations with a job array and multithreading on a supercomputing cluster.
old_optimizer contains the older python implementation, qevo_ntom.py, which calculates exact fidelities and has a few different features. The exact calculations come with an exponential increase in computation with the number of registers. Example.ipynb runs through an example purifying to two pairs.
- in calculate_performance!, first run the circuit with only initialization noise for fewer number of simulations and skip running the simulations with noisy operations if the fidelity is too low
- increasing the number of simulations as the generation increases