/Korg.py

For people who are not ready to switch from I̶D̶L̶ ̶t̶o̶ ̶P̶y̶t̶h̶o̶n̶ Python to Julia

Korg.py

This provides instructions on how to execute Korg.jl, a 1D LTE stellar spectral synthesis code for FGK stars in Julia, from Python.

Code paper: Wheeler et al. (2023). Please cite this if you use Korg.

Tutorial

Documentation

Installation

The following code snippet will install and link Julia 1.9.1 to your current Python environment, and install the latest version of Korg.jl.

wget https://raw.githubusercontent.com/abelsiqueira/jill/v0.4.0/jill.sh
sudo bash jill.sh -y -v 1.9.1
export PYTHON=$(which python)
julia -e 'using Pkg; Pkg.add("PyCall")'
pip install julia
python -c 'import julia; julia.install()'
unset PYTHON
julia -e 'using Pkg; Pkg.add("Korg")'

Example

This repository contains the example line list and model atmosphere necessary to execute the example below.

import matplotlib.pyplot as plt
from julia import Korg

lines = Korg.read_linelist("linelist.vald", format="vald")
atm = Korg.read_model_atmosphere("s6000_g+1.0_m0.5_t05_st_z+0.00_a+0.00_c+0.00_n+0.00_o+0.00_r+0.00_s+0.00.mod")
A_X = Korg.format_A_X(0)
sol = Korg.synthesize(atm, lines, A_X, 5160, 5176)

fig, ax = plt.subplots(figsize=(12, 4))
ax.plot(sol.wavelengths, sol.flux, "k-")
ax.set_xlabel("$\lambda$ [Å]")
ax.set_ylabel("$F_\lambda/R_\mathrm{star}^2$ [erg s$^{-1}$ cm$^{-5}$]")
fig.savefig("example.png")

example

FAQ / Issues

I have some kind of linking errors.

Try python-jl instead of python, or python-jl -m IPython instead of ipython. Can also run this within python/ipython before doing anything: from julia.api import Julia; jl = Julia(compiled_modules=False)

I am having trouble installing on an ARM (M1/M2) Mac with anaconda

  • This doesn't yet work with ARM, so change it to use x86. Create a new environment without installing anything, then run:

conda config --env --set subdir osx-64

This makes your environment install with osx-64. Once that is done, then install python. To get the example to work:

conda create -n korg
conda activate korg
conda config --env --set subdir osx-64
conda install conda install python matplotlib scipy astropy numpy jupyter ipython

Then in ipython:

from julia.api import Julia; jl = Julia(compiled_modules=False) # takes a moment

import matplotlib.pyplot as plt
from julia import Korg

lines = Korg.read_linelist("linelist.vald", format="vald")
atm = Korg.read_model_atmosphere("s6000_g+1.0_m0.5_t05_st_z+0.00_a+0.00_c+0.00_n+0.00_o+0.00_r+0.00_s+0.00.mod")
A_X = Korg.format_A_X(0)
sol = Korg.synthesize(atm, lines, A_X, 5160, 5176)

fig, ax = plt.subplots(figsize=(12, 4))
ax.plot(sol.wavelengths, sol.flux, "k-")
ax.set_xlabel("$\lambda$ [Å]")
ax.set_ylabel("$F_\lambda/R_\mathrm{star}^2$ [erg s$^{-1}$ cm$^{-5}$]")
fig.savefig("example.png")

Recall that Korg will take a while to run the first spectrum, because it needs to compile. But if you run it again, it will be very fast each subsequent time, so you win when making a large number of spectra.