Data structures
Astrocook manages three main data structures: spectra, line lists, and system lists. All three structures include a data table (actually an Astropy Table object) and a metadata dictionary. These structures and other ancillary structures are bundled into a .acs archive when you save a snapshot of a session.
Table of contents
Spectra
Each Astrocook session is based on a spectrum, i.e. a table that pairs a flux-like quantity to a wavelength-like quantity. We use the expressions “flux-like” and “wavelength-like” to be as general as possible: from the point of view of structure, it doesn’t matter if the spectrum is expressed in wavelengths or frequencies, or if the flux density is calibrated in physical units or not.
Spectrum tables are saved both in FITS and ASCII format, with file names snapshot_spec.fits/.dat respectively (assuming snapshot.acs is the name of the archive containing them).
If the continuum level of the flux has been determined, a table of continuum nodes may be present, depending on how the continuum was determined (see here for more details). The node tables are also saved in FITS and ASCII format, with file names snapshot_nodes.fits/.dat respectively.
List of lines
When you detect spectral lines (either in emission or absorption), they are formatted into a table that is added to the session.
Line tables are saved both in FITS and ASCII format, with file names snapshot_lines.fits/.dat respectively.
List of absorption systems
Absorption lines detected in quasar spectra are frequently grouped into absorption systems, i.e. sets of lines produced by different ions at the same redshift.
In our convention, a system has one and only one redshift. This means that e.g. different doublets at a similar redshift are treated as separated (one redshift per doublet). It also means that each component of an absorbers with a complex velocity structure (also sometimes called a “system”) is considered a system in itself. Here you can find a complete list of the ionic transitions used to model absorption features.
System tables are saved both in FITS and ASCII format, with file names snapshot_systs.fits/.dat respectively.
System tables only contain the parameters needed to compute the system models. The models themselves (i.e. the fitting functions, which often group together several systems into a single expression, when they need to be fitted together) are saved in a set of ASCII files snapshot_systs_mods_NN__lines.dat/__group.dat/_left.dat/_right.dat, with NN the ID of the model. An additional ASCII file snapshot_systs_mods.dat provides a list of the models. These files are handled internally via serialization by pickle and lmfit.save_model()/.load_model() and are not meant to be accessed by the users.