Installing

1. Clone this repo using git clone git@github.com:jaynavar/Copysets.git.
2. cd into the Copysets repo.
3. Install the dependencies using sudo pip install -r requirements.txt.

Running Figure 6 reproduction

$ python2 Main_Figure6.py [options]

Use -h to list all options available. Some important parameters are -r which allows you to render the graph locally without saving to a file, and -s which saves all data and figures to a timestamped directory under data_Figure6. You can use the --simulation flag to generate the data using simulations rather than the equations.

Running Repeated Failures experiment

$ python2 Main_RepeatedFailures.py [options]

Use -h to list all options available. Some important parameters are -r which allows you to render the graph locally without saving to a file, and -s which saves all data and figures to a timestamped directory under data_RepeatedFailures.

Viewing Experiment Data

All data from the experiments used in the report are stored under data_Figure6 and data_RepeatedFailures. Each experiment includes a TRIAL_INFO.txt file which lists the parameters used for the given experiment. You can reproduce the figures from any given experiment by running the appropriate tool (i.e. Main_Figure6.py or Main_RepeatedFailures.py) with the --load <experiment_dir>/DATA.json option. Note that for Main_Figure6.py you must ensure the parameters passed into the tool are the same as those specified in the TRIAL_INFO.txt file for the given experiment. The data files are stored in the experiment directories in a file labelled DATA.json.