DBNets2.0

Dust Busters Nets 2.0 - Simulation-based inference pipeline for characterizing disc substructures and putative embedded planets. Unlike the first version this is only offered as a python library. DBNets2.0 introduces a powerful upgrade: it simultaneously fits the putative planet mass along with three additional disc properties that can degenerately lead to similar substructures.

Check out the paper here: DOI.10.1051/0004-6361/202554401

To install this library

1. open a terminal

2. clone the repository with git clone https://github.com/dust-busters/DBNets.git

3. enter the new directory with cd DBNets

4. install the library with pip install .

5. download the trained models using the provided script dbnets-download. It should be available from the terminal after the package is correctly installed.

If you encounter some errors following the previous instructions, you can try to install the package in a python enviroment. To do that, you can follow the instructions below.

Install in a virtual enviroment

1. First create a new python enviroment with python3.10 -m venv <env_name>

2. activate the new enviroment source <env_name>/bin/activate

3. follow the above instructions to install DBNets in the new python enviroment

4. Enjoy!

To use the new enviroment within a jupyter-notebook, for instance for running the examples provided, create a new jupyter kernel using

python -m ipykernel install --name=<env_name>.

Once this is done, it is possible to select the new kernel from any jupyter-notebook.

Tutorials

There are some tutorial notebooks available in . The notebook contains a basic usage example of DBNets2.0.

Stay tuned for more documentation and examples, or drop me an email!