Python library for mocking spectral observations of galaxies

Originally conceived to mimic the PFS instrument at SUBARU, PFS_spectra can be easily adapted to any other configuration.

Juan Carlos Basto Pineda
Jérémy Dumoulin

To have a quick first idea of the way to use PFS_spectra just open the file example.params.config provided in the main distribution.

As you see there, this file is used to feed PFS_spectra all the necessary parameters regarding the instrumental and observational set up to be mimicked.

The params.config file is divided in three/four sections.

1. [geometry] aggregates the parameters related to the geometry of the mock observation.
2. [models] aggregates the parameters needed to define the models for the intensity and velocity maps.
3. [instrument] aggregates the parameters relevant to define the exact instrumental configuration to be mocked.
4. [environment] is projected (but not used yet) to store the variables related to the filesystem/environment.

Once you put all the required parameters in the ConfigFile, the way to pass them to PFS_spectra is with the functions stored in the file read_params.py. The main function there is get_allinput. As you see, it receives as input the ConfigFile, reads in the variables stored therein, and create three objects which will have those values as their attributes. Each object correspond to one of the blocks described before, as identified by the names of the classes and of the objects themselves:

geom    = GeometryObj()
geom.parse_input(ConfigFile)
models  = ModelsObj()
models.parse_input(ConfigFile)
instr   = InstrumentObj()
instr.parse_input(ConfigFile)

For instance, across the code the geometrical parameters can be invoked at any time through calls of the instances geom.incl, geom.pa, geom.z, etc. To get the exact name of all the instances please refer to the read_params.py file.

OK, so far PFS_spectra has read the input parameters and created the objects geom, models, and instr enclosing all the relevant information. Now, how to use it to mock the observations?

The first step is to create the mock velocity and intensity maps. To do that, you simply need to use the function make_velmap(geom, models, instr) from the velocity_models.py module, and the function make_fluxmap(geom, models, instr) from the flux_model.py module. Those functions make use of the routine sky_coord_to_galactic in the module tools.py, to project the coordinates in the plane of the disc to the coordinates in the plane of the sky.

With the maps at hand, the main routines for the next step are in the file operations.py. This includes a function to parametrize the emission line coming out of each pixel of the maps, shifted as necessary according to the cosmology and to the proper line-of-sight velocities. A function to record the footprint of the fiber convolved with the psf, which allows to weight the contribution of the flux of individual pixels in the final spectrum, and the main function for the creation of mock spectra: make_spectrum(instr, flux, lambda_obs, sigma, chunk = 50).

To understand the order in which these functions should be concatenated in a clean and efficient way, please check the example test_0.py, provided in the repository. A second example is given in test_1.py. After that you are invited to navigate through the code to get to know better the details.

As a final remark, just a friendly reminder for python beginners:

In order for PFS_spectra to work, you need to make its routines accessible to python from wherever you are working. This can be achieved en either of 3 ways:

1. by copying all necessary files in the working directory
2. by adding the location of the repository to the sys.path variable at the beginning of your script with:

import sys
sys.path.append("path_to_PFS_files")

3. By adding the location of the repository permanently to your PYTHONPATH variable, adding at your .bashrc file something like:

export PYTHONPATH="${PYTHONPATH}:/home/juan/Desktop/PFS_spectra/"

Then you can import the necessary modules, doing for example:

import read_params as rp
import operations as op