Run the demo experiment#

A good way to familiarize yourself with the code base and how to use it is to run the small demo experiment that we are providing in the /hsr4hci/experiments/demo folder. It is essentially a simplified version of Experiment 5.1 from our paper that uses a smaller temporal grid and a heavily binned version of our Beta Pictoris L’ data set to make sure it runs quickly even on a single machine.

To run the demo experiment, you will need to perform the following steps:

  1. Before you get started, make sure that hsr4hci and its dependencies are installed and properly set up. Check the Getting started guide for more information.

  2. If you have not done so already, download the data sets and place them in the right location. Check the guide about Data sets for more information.

  3. Now, to start and run the experiment, use:

    python $HSR4HCI_SCRIPTS_DIR/experiments/single-script/01_run_pipeline.py \
  --experiment-dir $HSR4HCI_EXPERIMENTS_DIR/demo

    See Getting started for more information about the environmental variables.

  4. Depending on your hardware, the experiment should take about 10 to 20 minutes to run. If you find this too long, you can, for example, open the config.json in the demo folder and try to increase the binning_factor further.

  5. Once they experiment has finished, you can find the results in $HSR4HCI_EXPERIMENTS_DIR/demo/results. The signal estimate is stored as a FITS file, which you can view, for example, using SAOImageDS9.

  6. You can also generate a plot as a PDF by calling the following script:

    python $HSR4HCI_SCRIPTS_DIR/experiments/evaluate-and-plot/evaluate_and_plot_signal_estimate.py \
  --experiment-dir $HSR4HCI_EXPERIMENTS_DIR/demo

    You will find the final result plot in $HSR4HCI_EXPERIMENTS_DIR/demo/plots.

Feel free to use this as a starting point to explore our code in more detail. We have tried our best to document and comment everything, and we hope it should not be too hard to find your way around.