An automated PyRAF data reduction pipeline for GMOS spectroscopic data
The best way to install all the requirements (for both IRAF and Python) is to follow the instructions on the Gemini website. Note that IRAF/PyRAF require Python 2.7 rather than the generally recommended Python 3.x.
Download the latest (stable) version from this repository. Then run:
python setup.py install [--user]
where the --user flag is recommended so as not to install as root. To check whether the pygmos executable is in your PATH, type
which pygmos
If you see nothing (or a 'not found' message), it means the location of the executable is not part of your PATH. If this is the case, look through the installation messages for the place where the executable was copied to. In my case:
Installing pygmos script to /u/sifon/.local/bin
You should therefore add the equivalent of /u/sifon/.local/bin to your PATH. Add the following line to your ~/.bashrc or ~/.bash_profile:
export PATH=/u/sifon/.local/bin:$PATH
or to the ~/.cshrc, etc, if you're not using bash:
setenv PATH /u/sifon/.local/bin:$PATH
and restart the console.
First, make sure you activate the conda environment (let's assume it's called gemini):
source activate gemini
This should be done every time a new shell session is started. After this, the code is ready to do its magic:
pygmos <object> [options]
To see available options, type
pygmos -h
pygmos can also be used to display a GMOS mask using matplotlib, either by creating a matplotlib.collections.PatchCollection object or with aplpy.FITSFigure.show_regions:
from pygmos.mask import Mask
mask = Mask(mdf_filename)
# returns a matplotlib.collections.PatchCollection object
slit_collection = mask.collection()
# returns a region file name to be passed to aplpy.FITSFigure.show_regions
regfile = mask.regions()How it works:
The pipeline takes the object name given in the command line and finds all data associated with that object. It bias-subtracts all images and calibrates the science image with the flat field. It then does the wavelength calibration, removes cosmic rays using L.A.Cosmic (van Dokkum, 2001, PASP, 113, 1420; distributed with permission) and sky subtracts the spectra. After this, the individual exposures are added. Finally, the 1d spectra are extracted.
Data format taken by pygmos:
Usual GMOS FITS files, which means a set of exposures per object, each of which is composed of a science image, a flat field and a calibration arc.
When running the pipeline, keep in mind that:
- As of now, the pipeline reduces both MOS and longslit GMOS spectra but flux calibration is not implemented.
- Being an automated process, some things could go wrong. Most task parameters have to be modified by digging into the code, although some of the most important are easy to find, as they are in the definition of the functions. Others can be easily added in the usual IRAF way.
- This code has only been tested (and is recommended) for redshift measurements.
- Automatic identification of all relevant files given the object name.
- Incorporates the Lagrangian Cosmic Ray Removal "L.A.Cosmic" code implemented by P. van Dokkum (distributed with permission).
- Can be executed either in automatic or interactive mode, which allows for a more thorough analysis, without the need to run each PyRAF task separately.
- Has the option of automatically cutting the spectra and copying them to a separate folder. This is useful if, for instance, the spectra will be cross-correlated using RVSAO, which takes only single spectra (as opposed to multi-slit images) as input.
- Has the option of aligning the 2d images, which is particularly useful for visualizing galaxy cluster data.
- Flux calibration is not implemented.
- There is no automated search for the bias file(s). The (master) bias file needs to be given in the command (see help page). If not given, the pipeline will ask for one.
- The inventory only finds one Flat and one Arc per observation.
- When run automatically, the pipeline only extracts one aperture from each slit, while some slits might contain more than one object.
- The interactive feature runs all tasks interactively, without being able to switch individual tasks as automatic and others as interactive.
(c) Cristóbal Sifón
Last updated October 2017.