oc-py: A Modular Python Suite for Eclipse Timing Variation Analysis

oc-py is an open-source, platform-independent Python-3 library designed for high-precision modeling and analysis of Eclipse Timing Variations (ETV). Developed with a modular architecture, the software provides a robust pipeline from raw data processing to complex modeling and Bayesian inference.

By leveraging standard scientific libraries, oc-py ensures reliability and reproducibility in modeling orbital period variations, whether they are periodic or cyclic driven by light-travel time effect (LiTE), dynamical perturbations, or stellar magnetic activity; or secular driven by mechanisms such as mass transfer / loss, gravitational radiation.

Key Features & Modules

oc-py is structured into functional modules to facilitate a flexible research workflow:

  • Data Pre-processing: - Read data: Multi-format support (ASCII, CSV, Excel) via pandas and astropy. - Time conversion: High-precision conversion of diverse time scales to BJD-TDB. - Weighting and binning data: Statistical weighting (by instrument or uncertainty) and seasonal data folding. - Outlier removal: Automated identification and removal of discrepant data points using robust statistical methods.

  • Modeling & Fitting: - Analytical Models: Implements linear, quadratic (orbital decay), and Keplerian (LiTE) models based on Irwin (1952) using lmfit and scipy. - Dynamical N-Body Integration: Advanced Newtonian modeling of multi-body systems using the REBOUND integrator. - Stellar Activity: Models for magnetic activity-induced variations based on Applegate (1992), Völschow (2016), and Lanza (2020). - Orbital Decay & Growth: Quadratic functions to model secular orbital period variations. - User-Defined Models: A flexible interface for integrating custom analytical ETV functions combining different effects.

  • Statistical Analysis: - MCMC Sampling: Full Bayesian parameter estimation and posterior probability mapping using PyMC. - Model Comparison: Statistical model selection (AIC, BIC, \(\chi^2_\nu\)) to distinguish between competing models.

  • Visualization & Output: - Plotting: Graphics and plots for publication-ready O-C curves, and corner plots (parameter correlations). - Output parameters: Automated generation of formatted tables (LaTeX, CSV) for direct inclusion in scientific papers.