Spectractor

The goal of Spectractor is to measure the atmospheric transmission and intempt extracting spectra from slitless spectrophotometric images. It has been optimized on CTIO images but can be configured to analyse any kind of slitless data that contains the order 0 and the order 1 of a spectrum. In particular it can be used to estimate the atmospheric transmission of the LSST site using the dedicated Auxiliary Telescope.

Spectractor is structured in three subpackages:

• spectractor.extractor: extracts as most information as possible from a slitless data image, as the amplitude of the spectrum, the PSF evolution with the wavelength, the pixel to wavelength calibration, an estimate of the background under the spectrum, the position of the order 0;
• spectractor.simulation: contains all the tools to simulate a spectrogram, as the atmospheric transmission simulation, the inclusion of instrumental throughput, the simulation of mock slitless data images;
• spectractor.fit: compares the extracted data with simulations to estimate the atmospheric transmission and refine the spectral extraction.

Some submodules complete the structures with generic functions:

• spectractor.parameters: contains all the global parameters of Spectractor to set its general behaviour, the instrumental characteritics, etc;
• spectractor.config: tools to read config .ini text files and set the global parameters;
• spectractor.logbook: tools to read logbook .csv text files and get some metadata relative to the data images that are not contained in the header;
• spectractor.tools: contains generic functions shared by all the subpackages (fitting procedures, plotting functions, etc).

Installation

Spectractor is written in Python 3.9. The mandatory dependencies are listed in the requirements.txt file. To install Spectractor, just run

pip install -r requirements.txt .

To simulate atmospheric transmission, Spectractor needs libradtran (optional). There are two ways to install this software:

• use the Rubin observatory conda package

conda install -c conda-forge rubin-libradtran

• compile it from sources: it needs the installation of netcdf and a python 2 environment (for the compilation only, not the usage); uvpsec executable must in the user $PATH or the user has to set an environmental variable $LIBRADTRAN_DIR pointing to the install directory.

To find accurate star centroids using astrometry, one can install astrometry.net (optional). There are two ways to install this software:

• use the anaconda package

conda install -c conda-forge astrometry

and then download and move index files in the $CONDA_PREFIX/data folder.

• compile it from sources: it needs the installation of netpbm and wcslib; solve-field executable must in the user $PATH or the user has to set an environmental variable $ASTROMETRYNET_DIR pointing to the install directory.

For faster matrix multiplications, if available for a given computer configuration, one can install sparse_dot_mkl:

conda install -c conda-forge sparse_dot_mkl

Spectractor is able to perform parameter fits using the MCMC library emcee (optional) with mpi4py and h5py. The latter might be better installed using conda install ... command to get their own dependencies (openmp and hdf5).

Basic extraction

The main file is spectractor/extractor/extractor.py with the function Spectractor. It extracts the spectrum from a science data image (deflatted, debiased), given:

• the path to the FITS image from which to extract the image,
• the path of the output directory to save the extracted spectrum (created automatically if it does not exist yet),
• the rough pr exact position of the object in the image (in pixels),
• the name of the disperser (as it is named in the spectractor/extractor/dispersers/ folder),
• the name of the config .ini file,
• optionally the name of the target (to search for the extra-atmospheric spectrum if available).

filename="./tests/data/reduc_20170530_134.fits"
output_directory="./outputs/"
guess = [745,643]
disperser_label = "HoloAmAg"
config = "./config/ctio.ini"
target = "HD111980"

Then the spectrum is simply extracted from the image and saved in a new fits file using the Spectractor function:

spectrum = Spectractor(filename, output_directory, guess=guess, target_label=target, disperser_label=disperser_label, config=config)

or typing the following command within a terminal:

python runExtractor.py ./tests/data/reduc_20170530_134.fits -o outputs --config ./config/ctio.ini --xy [745, 643] --target HD111980 --grating HoloAmAg

Spectractor comes with two verbosity modes, set in the parameters.py file:

• VERBOSE (or -v, --verbose) : the first level of verbosity that returns many information along the process to know what the program is doing; it also plots the output spectrum
• DEBUG (or -d, --debug) : in the debugging mode some intermediate plots are produced to see the performance of the program.

Plot a spectrum

To see the result of the extraction process, it is possible to load it via the Spectrum class and plot it:

spectrum = Spectrum('./tests/data/reduc_20170530_134_spectrum.fits', config="./config/ctio.ini)
spectrum.plot_spectrum()
spectrum.plot_spectrogram()

This object is also returned by the Spectractor function.

Tutorial notebook

A tutorial Jupyter notebook is available in the notebooks folder.