This repository holds the code and data used in Optimizing spectroscopic follow-up strategies for supernova photometric classification with active learning, by Ishida, Beck, Gonzalez-Gaitan, de Souza, Krone-Martins, Barrett, Kennamer, Vilalta, Burgess, Quint, Vitorelli, Mahabal and Gangler, 2018.

This is one of the products of COIN Residence Program #4, which took place in August/2017 in Clermont-Ferrand (France).

We kindly ask you to include the full citation if you use this material in your research: Ishida et al., 2018 - arXiv:astro-ph/1804.03765.

The examples folder contains all the necessary steps to perform the basic steps in the pipeline:

1. Light cure fit using the parametric function from Bazin et al., 2009
2. Separate the photometric sample in query/target
3. Get the simulated features used to identify neighbors
4. Build pseudo-training sample for the canonical strategy
5. Run the active learning algorithm - we use the Active Learning implementation in libact

The current version runs in Python-2.7.

In order to install this code you should clone this repository and do:

    python setup.py install --user

- Python=2.7
- numpy>=1.8.2
- matplotlib>=1.3.1
- pandas>=0.19.2
- scipy>=0.18.1
- astropy>=2.0.1
- libact>=0.1.3

You only need to install this package if you intend to use the SNANA file handling as shown in the examples folder.

If you already have your data reduced you can skip this part.

snclass