box-least-squares

The Box Least Squares (BLS) periodogram skill is a specialized data analysis resource designed for astronomers and researchers working with photometric time series data. It serves as an automated engine for detecting periodic signals hidden within light curves, specifically targeting the box-shaped transit signatures characteristic of exoplanets passing in front of their host stars or binary star systems eclipsing one another. By modeling the transit signal as a periodic upside-down top hat, the skill allows users to extract precise orbital parameters from noisy observational data, offering a robust alternative to other periodogram methods like Transit Least Squares (TLS) or Lomb-Scargle.

• Performs automated transit detection using the astropy.timeseries library, facilitating integration into broader astronomical data processing pipelines.

• Calculates critical transit variables including orbital period, transit duration, flux depth, and mid-transit reference time (T0).

• Supports flexible objective functions, allowing users to switch between maximum likelihood estimation and Signal-to-Noise Ratio (SNR) optimization to better handle correlated noise in observations.

• Includes advanced statistical validation methods such as odd-even transit depth comparison, which helps distinguish between planetary transits and eclipsing binary false positives.

• Provides automated period grid generation via the autopower method, ensuring that period searches are both efficient and computationally thorough.

• Input: Requires time-series photometric data, typically provided as arrays of time, flux, and optional flux error values; requires astropy units for consistent temporal and flux scaling.

• Output: Returns a periodogram object containing arrays of periods, power values, durations, and transit times, along with stats dictionaries for specific candidate peaks.

• Usage constraints: Users must ensure high-quality period grid spacing to avoid missing the true period; significantly coarse grids or improper frequency ranges may lead to alias identification or missed signals.

• Practical tips: Always perform validation using compute_stats to analyze the odd-even mismatch; high SNR values (>7) and consistent depths across multiple transits are key indicators of a viable candidate.

• Integration: Best used as a module within larger analysis workflows for planetary discovery, following initial data reduction and cleaning procedures typical in astronomical research.