Reproduction package

Code and artifacts accompanying the paper

@inproceedings{Schoenberger.2021,
  author    = {Schönberger, Manuel and Franz, Maja and Scherzinger, Stefanie and Mauerer, Wolfgang},
  title     = {Peel | Pile? {Cross}-Framework Portability of Quantum Software},
  booktitle = {19th IEEE International Conference on Software Architecture (ICSA)},
  address   = {Honolulu, HI, USA},
  publisher = {IEEE},
  year      = {2022}
}

Docker

Get docker image

Build image:

docker build -t qsa-repro .

or pull image:

docker pull ghcr.io/lfd/qsa-repro/qsa-repro:latest

Create Container

docker run --name qsa-repro -it qsa-repro [<-flags>] [<option>]

The <option> specifies which operations are performed on container start.

Available options are:

• experiments_only: performs the operations of the experimental Analysis (RL trainings and MQOs)*
• rl_only: performs RL trainings*
• mqo_only: performs MQO experiments
• paper_only: generates the full paper from LaTeX
• all: performs all of the above*
• bash(default): does not perform any operation, but launches interactive shell

Feel free to define additional <-flags>, e.g.:

• Volume, to keep track of generated files on the host system: -v $PWD:/home/repro/qsa-repro
• Port forwarding to launch TensorBoard on the container to track the training process for RL on the host: -p 6006:6006. TensorBoard can be started in the Container with: tensorboard --logdir expAnalysis/RL/logs --host 0.0.0.0

*Please note the long runtimes for RL trainings (several days). For quickly inspecting our reproduction package, we recommend to use the options mqo_only, paper_only or bash.