Introduction
Usage
The current version of CAFE (v1.0.0) is released together with a jupyter notebook that walks the user through the process of fitting a spectrum obtained from JWST MIRI/MRS data. The jupyter notebook, named CAFE_tutorial_NGC7469_1D.ipynb, can be found in the notebooks/ folder in the GitHub repository, and can be copied to a working directory created by the user.
CAFE is able to read any spectrum provided by the user in a simple .dat or .txt file format containing a table with columns reporting: wavelength, flux density, and flux density uncertainty (header/title/comment lines must be preceeded by a #). The user is responsible to ensure that there are no NaNs in the data. Also, the spectrum must be sorted in order of ascending wavelength and with no overlapping entries. (The CRETA extraction tool, which can be used to produce output spectra that can be fed directly to the CAFE fitter, exists in the code and takes care of these details, but is not currently supported in this first release).
The jupyter notebook starts by setting up the necessary paths pointing to where the data are located and where the output results from the fitting will be stored. Next, a spectrum of the galaxy NGC 7469 (Lai et al. 2022, ApJ, 941, 36) is downloaded from the cloud. After, the redshift of the source is set, together with the name of the file containing the spectrum, and the .ini and .cafe files necessary for the initialization of the fit (they are located in the CAFE/inp_parfiles/ and CAFE/opt_parfiles/ folders). These two last files control, respectively: (1) the initial value of the parameters and fit variables for each continuum component (continua, lines, PAHs, user opacities); (2) code-specific variables (we do not recommend changing any of these, as some have not been tested). We refer the user to the Spectral Fitting page for more details.
The spectrum can be plotted together with the initial (default) model decomposition for visual inspection. The user can tweak the initial model by modifying keywords in the .ini parameter file. These keywords refer either to model parameters themselves (e.g., peak line fluxes, dust component temperatures, etc.), constraints for each or a combination of model parameters (e.g., line width variation limits, optical depth variation limits for dust components, etc.), specific names of files, or general info needed for the setup.
Once the user is satisfied with the initial guess model, the spectral fitting can be run. CAFE uses the LMFIT python package to minimize the data-model residuals using the Trust Region Reflective least-squares method (least_squares), and based on the χ2 statistic.
All this steps can be found in the jupyter notebook inside the notebooks/ folder in the GitHub repository.
Output
The CAFE fitter returns a parameter (LMFIT) object containing the best/optimized parameters from which physical quantities (e.g., temperatures) and observables (e.g. line centroids) can be extracted or constructed (e.g., fluxes, based on the line centroid and width). All the parameter information is dumped into python dictionaries for further use and stored in data tables on disk, in the cafe_output/ directory. In addition, the parameter object is saved to disk as a .fits file in a ‘parameter cube‘, which can be read at any later stage to run further fits or generate new dictionaries or data tables (see the jupyter notebook section “RESTORE CAFE SESSION FROM DISK”). The optputs are stored in the cafe_output/ folder using a default name that is the same as the input spectrum file. In summary the output files from CAFE are:
_cafefit.asdf : A file containing dictionaries with all the parameter and component information, including the fitted parameters and the continuum profiles. The user can read this file from disk and re-plot the
CAFEfit or re-write the tables containing the fluxes of the features (lines and PAHs).
_parcube.fits : A file containing all the info necessary to recover a previous CAFE fitting session (see the jupyter notebook section “RESTORE CAFE SESSION FROM DISK”). This is useful if the user wants to refine the fit (for instance by assuming as initial conditions the results from the fit of another source; or by the result from the previous fit of the same source itself).
_fitpars.ini : A file containing all the fitted parameters in the format of an .ini file, which can be used to initialize the fit of any other spectra (similar to the parcube.fits option). It also serves for book-keeping; i.e., to keep track of the files used to define the emission lines to be fitted via the lines.H.recombination_***.txt, lines.atomic_***.txt and lines.molecular_***.txt files.
_fitfigure.png : A file containing a figure of the last fit performed.
_linetable_*.ecsv : A file containing the * (=int: intrinsic; =obs: observed) fluxes of the fitted emission lines.
_pahtable_*.ecsv : A file containing the * (=int: intrinsic; =obs: observed) fluxes of the fitted PAH features.