BLiSS — Blind Line Search System

BLiSS (Blind Line Search System) is an open-source Python package for finding emission lines in spectra without telling the code where to look first.

In other words: a real blind line-search algorithm — you only need your spectrum.

BLiSS was designed to simplify and speed up spectral analysis in X-ray astronomy. Instead of manually inspecting spectra, guessing candidate energies, and testing lines one by one, BLiSS performs an automatic search for statistically significant emission features and then cross-matches them with atomic transitions.

The package combines Gaussian line-profile fitting, statistical scoring, and database-based identification using XSTAR atomic data. It is especially useful when the continuum is complex, uncertain, or simply not the main focus of the analysis.

Although BLiSS was developed for X-ray spectra, the core idea is more general: if you have spectral data with uncertainties, BLiSS can help you search for emission features in a systematic and reproducible way.

Main features

BLiSS is designed to make emission-line searches fast, systematic, and reproducible. Its main features include:

  • Real blind emission-line search
    No prior list of candidate energies is required. BLiSS searches the spectrum directly and identifies statistically relevant emission-like features.

  • Automatic Gaussian fitting
    Candidate features are modeled with Gaussian profiles and ranked using statistical scores, helping users decide which lines deserve further inspection.

  • Continuum-independent workflow
    BLiSS does not require a predefined global continuum model to start searching for lines, making it especially useful for complex or uncertain spectral shapes.

  • Atomic line identification with XSTAR
    Detected candidates can be cross-matched with atomic transitions from XSTAR, including Doppler-velocity constraints to account for shifted lines.

  • Flexible rebinning utilities
    BLiSS includes tools to rebin spectra according to instrumental resolution or to reach a target signal-to-noise ratio.

  • ISIS integration
    Detected emission lines can be automatically added to ISIS spectral models, simplifying follow-up fitting and interpretation.

  • Reproducible analysis
    The same search criteria can be applied consistently across many spectra, reducing subjective choices and making BLiSS useful for large datasets.

If you have any questions or need assistance, please feel free to reach out: graciela.sanjurjo@ua.es.

Getting started

Installation

Here you can find the source code.

License

MIT license