Docs/API/ExoParams

ExoParams

Convenience container for model input features in a clear, named form. Missing values can be left as None; they will be imputed before prediction/training.

Import

Bring in the ExoParams container when you want named, self-documenting inputs.

from exobengal.exobengal import ExoParams

Constructor

All fields are optional; None will be imputed and scaled during preprocessing.

ExoParams(
  period=None,   # koi_period (days)
  prad=None,     # koi_prad (Earth radii)
  teq=None,      # koi_teq (K)
  srad=None,     # koi_srad (Solar radii)
  slog_g=None,   # koi_slogg (log10 cm/s^2)
  steff=None,    # koi_steff (K)
  impact=None,   # koi_impact
  duration=None, # koi_duration (hours)
  depth=None,    # koi_depth (ppm)
)

Usage

This form improves clarity and reduces ordering mistakes compared to raw lists.

from exobengal.exobengal import DetectExoplanet, ExoParams

params = ExoParams(period=365.0, prad=1.0, teq=288.0, srad=1.0, slog_g=4.44, steff=5778, impact=0.1, duration=5.0, depth=100.0)
detector = DetectExoplanet()
result = detector.random_forest(params)

Feature Order

The underlying models expect features in this exact order when passing a list:

[period, prad, teq, srad, slog_g, steff, impact, duration, depth]

DetectExoplanet

Utilities