Person responsible for data: Adam Glos (aglos [at] iitis.pl).
The scripts necessary for generating the results provided in the "Error mitigation for variational quantum algorithms through mid-circuit measurements"
The code was used on Ubuntu OS 20.04. It is not guarantee it will work on other operating systems
Anaconda distribution can be downloaded from https://www.anaconda.com/products/individual. To install
conda env create -f hobo_encoding_change.yml
To activate this environment, use
conda activate hobo_encoding_change
To deactivate an active environment, use
conda deactivate
Julia can be downloaded from the official website https://julialang.org/. Version 1.6.1 was used.
In order to set up the environment, please go to the directory were qaoa_efficiency_analysis. Then run julia and use
julia> ]
(@v1.6) activate .
(@v1.6) instantiate
This will install the required packages based on Manifest.toml file. The environment will be activated on its own when running scripts.
The repository already contains the data used in the publications. To generate new samples, please follow the instruction below. Occasionally the code omits generating new data if old already exist (the scripts will complain, or inform about such issue) - in such case please remove the old data.
To generate data, run the following files in the qaoa_efficiency_analysis directory:
julia tsp_generator.jl tsp_data_xy/tsp3 3 100
julia tsp_generator.jl tsp_data_xy/tsp4 4 100
This will generate 100 TSP for 3 and 4 cities located at the corresponding repositories.
To generate data for random angles for XY-QAOA and GM-QAOA, please run in the main directory
./xy_experiment_generator.sh
This will generate data of appropriate size used in the paper. To parallelize the compution, in the bash script we used & so that the next command is run after the previous was started. This will likely produce large number of processes for a regular PC. To surpass overloading the computer, please remove &.
The commands run through this script takes form
python energy_diff_generator.py $VAR $CITYNO 40 1 100 rand $GAMMA
where VAR is the type of error model used for computation, CITYNO is the number of cities, 40 is the number of layers considered, 1 is the number of times a single TSP instances should be considered 100 is the number of different TSP instances, rand is the type of angles used (the only option working) and GAMMA is the number of noise strength chosen.
Furthermore. the script runs
python optimization_qaoa.py 3 40
python optimization_qaoa.py 4 40
which runs the QAOA algorithm for 3/4 cities and 40 TSP instances (each considered once). To reduce the number of parallel processes, please change the pool_size in the file optimization_qaoa.py.
For Energy difference computation, data will be saved to data_noise_qaoa_xy_gm/tsp3_plotting_for_paper and data_noise_qaoa_xy_gm/tsp4_plotting_for_paper depending on the number of cities considered. For QAOA optimization, the data will be saved to data_noise_qaoa_xy_gm/optimization.
To generate plots, please run the following commands
python energy_diff_plot.py 40 1 100 rand
python prob_diff_plot.py 40 1 100 rand
python optimization_plot.py
The last command may require slightly more time than the previous one, as the energies has to be recomputed. The plots will be saved to plots.