Rad1D

A simple 1D, plane-parallel radiative transfer code that implements accelerated lambda iteration (ALI).

This project was made in a collaboration including Manuel Barrientos, Anthony Burrow, Adam Moss, and Sarah Stangl.

For more details on this project, see .doc/FinalProjectReport.pdf.

Requirements

This may be compiled and run on either Windows or Linux platforms with an available C++ compiler. To use Rad1D as a Python-wrapped module, the following are required:

• Python 3.6+
• pybind11 package

Installing the code

Rad1D may be installed with pip (setuptools) by:

git clone https://github.com/anthonyburrow/Rad1D.git
pip install ./Rad1D/

Running with Python

The RadModel class is exposed to Python and instantiated in the following way:

from Rad1D import RadModel

params = {
    # Parameter key-values here
}

model = RadModel(params)

See "./examples/example1.py" for how RadModel may be used. Remember to set all absolute paths in the global parameters with the correct paths for the user machine.

Parameters

Below are the currently available global parameters, and their default values:

params = {
    'data_dir'       : '../data',   # Absolute path to line list
    'wave_start'     : 4000.,       # Starting wavelength
    'wave_end'       : 7000.,       # Ending wavelength
    'cont_res'       : 0.05,        # Points per angstrom resolved for continuum
    'line_res'       : 3.0,         # Points per angstrom resolved for lines in line list
    'tau_min'        : 1e-8,        # Minimum tau of atmosphere for continuum
    'tau_max'        : 1e6,         # Maximum tau of atmosphere for continuum
    'eps'            : 1e-4,        # Thermalization factor
    'T_eff'          : 6000.,       # Characteristic temperature (K) of atmosphere
    'n_zones'        : 256,         # Number of tau points
    'max_iter'       : 500,         # Maximum number of lambda iterations allowed
    'accelerated'    : True,        # Use accelerated lambda iteration
    'Ng_accelerated' : True,        # Implement Ng acceleration
    'eps_converge'   : 1e-8,        # Factor to determine J is converged
    'n_quad'         : 32,          # Order of Gaussian quadrature integration
    'verbose'        : True,        # Display stdout output
}

Running the Tests

One may run the suite of tests that generates plots in the documentation with

pytest -vs ./Rad1D/tests

(Point to the tests directory accordingly.)