firepype 0.0.1

FIRE AB-pair NIR reduction pipeline

firepype

Python pipeline for Magellan/FIRE Prism-mode data. Experimental; validated on a limited dataset. Verify outputs against established pipelines (e.g. FireHose_v2).

Features

• Slit edge and object detection with robust heuristics
• Arc 1D extraction and line matching
• Robust Chebyshev dispersion solution with optional anchors
• Parity-aware A–B/B–A differencing and robust negative-beam scaling
• Footprint median extraction with error estimates
• Interpolation-edge masking and inverse-variance coaddition
• Telluric correction:
  • POS-only standard extraction
  • Band-wise scaling with deep-gap masking
  • Vega model broadened to the instrument resolution
  • Transmission (T) smoothing only inside dense contiguous regions
• Optional QA plots: arc overlays, labeled arc 1D, final coadd, telluric T(λ), corrected spectra

Installation

Installation (PyPi):

. .venv/bin/activate  # Windows: .venv\Scripts\activate
python -m pip install -U pip
pip install firepype

Alternatively, install from source:

. .venv/bin/activate
python -m pip install -U pip
git clone https://github.com/gemma-cheng/firepype
cd firepype
pip install -e .[dev]

Requirements:

• Python 3.10–3.12
• numpy ≥ 1.22, scipy ≥ 1.9, astropy ≥ 5.2, matplotlib ≥ 3.6

Quickstart

Minimal end-to-end reduction:

firepype \
  --raw-dir /path/to/raw \
  --out-dir ./out \
  --arc /path/to/raw/fire_0123.fits \
  --ref-list /path/to/ref/line_list.lst \
  --spec "1-4, 10-7"

Telluric/response only:

firepype-telluric \
  --standard ./out/standard_extracted.fits \
  --out-dir ./out/telluric \
  --stype A0V --plot

Outputs

• ./output/qa/ … if QA enabled
• Coadded spectrum FITS: wavelength_um, flux
• Telluric and response FITS in out/telluric/

Basic Tutorials

Basic usage tutorials can be found in the tutorials directory

Notes on telluric method

• Standard extraction: POS-only column, tuned aperture/background; alternative beam used if POS median is non-positive
• Wavecal: average across a small footprint around the chosen standard column, using ARC + line list
• Vega model: broadened to instrument resolution (R ~ 6000 by default) in log-λ space
• Continuum: robust Chebyshev fit to the standard/model ratio within each band (J/H/K), excluding deep telluric gaps and known A0V intrinsic lines; fit is used to normalize before deriving T
• Transmission T(λ): computed and lightly smoothed only within dense contiguous support (prevents spreading across gaps)
• Application: science flux and errors are divided by T within overlap where T_min ≤ T ≤ T_max; elsewhere, values are left as NaN to avoid artifacts

Limitations and validation

• Tested on a limited set of FIRE Prism-mode observations; results are not guaranteed.
• Validate outputs against established pipelines (e.g. FireHose_v2):
  • Wavelength RMS per region
  • Sky residuals around OH lines
  • Merged-order continuity
  • S/N consistency

License

MIT (see LICENSE).

File Structure

• firepype/
  • __init__.py — package API (exposes run_ab_pairs, apply_telluric_correction)
  • config.py — dataclasses for configuration
  • io.py — FITS I/O, path builders, ID pairing
  • utils.py — math helpers, masks, line-list loader
  • calibration.py — peak detection, line matching, dispersion solver
  • detection.py — slit/object detection, parity, negative scaling
  • extraction.py — 1D extraction routines
  • coadd.py — coaddition accumulator
  • plotting.py — optional QA plotting
  • pipeline.py — high-level AB-pair orchestration
  • telluric.py — telluric correction API
  • cli.py — command-line interface (pipeline + telluric subcommand if enabled)
• tests/ — minimal tests for core functionality
• pyproject.toml — packaging configuration
• README.md — this file
• LICENSE