Basic Example Lite

This example is the much condensed version of our basic example for those of us already familiar with the parameters and how LUCI works. Let’s get started!

Let’s start with importing LUCI. Just as descsribed on the readthedocs page, we need to point our system path to our install location of the LUCI folder. We will additionally import the Luci plotting functions :)

You can download the example data using the following command:

!wget -O M33.hdf5  https://ws.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/raven/files/cadc:CFHT/2309128z.hdf5

This will save a file called M33.hdf5 to the directory where you run this command. Do not be concerned if it takes a few minutes (or more than a few) to download.

The region files used in the examples can be found in the ‘Examples/regions’ folder. To run the examples, place these region files in the same directory as the hdf5 file.

# Imports
import sys
sys.path.insert(0, '/media/carterrhea/carterrhea/SIGNALS/LUCI/')  # Location of Luci
from LuciBase import Luci
import LUCI.LuciPlotting as lplt

Set the required parameters. We will be using our machine learning algorithm to get the initial guesses - this happens under the hood in LuciFit, so that the user is not required to think about the initial guess

# Initialize paths and set parameters
Luci_path = '/home/carterrhea/Documents/LUCI/'
cube_dir = '/export/home/carterrhea/M33'  # Path to data cube
cube_name = 'M33_SN3'  # don't add .hdf5 extension
object_name = 'M33'
filter_name = 'SN3'
redshift = -0.0006  # Redshift of object
resolution = 5000

We intialize our LUCI object

# Create Luci object
cube = Luci(Luci_path, cube_dir+'/'+cube_name, cube_dir, object_name, redshift, resolution)

The output will look something like this:

Let’s quickly create a deep frame

# Create Deep Image
cube.create_deep_image()

Let’s extract a background region and take a look at it. The background region is defined in a ds9 region file in the Examples/regions folder in the LUCI directory.

bkg_axis, bkg_sky = cube.extract_spectrum_region(Luci_path+'Examples/regions/bkg_M33.reg', mean=True)  # We use mean=True to take the mean of the emission in the region instead of the sum
lplt.plot_spectrum(bkg_axis, bkg_sky)

We now fit part of our cube defined by the bounding box 1200 < x < 1350 and 1700 < y < 1950 with a SincGauss on the Halpha line, the NII-doublet, and the SII-doublet with a binning of 2. We are also going to constrain our velocities and sigmas. We can also run multiple threads with the n_threads argument.

For this example, we do not calculate the errors because it slows down calculations, but note that it can easily be done by adding the argument uncertainty_bool=True. If you want the full Bayesian calculation you can add bayes_bool=True.

# Fit!
vel_map, broad_map, flux_map, ampls_map = cube.fit_cube(
    ['Halpha', 'NII6548', 'NII6583', 'SII6716', 'SII6731'],
    'sincgauss',
    [1,1,1,1,1], [1,1,1,1,1],
    1200, 1350,
    1700, 1950,
    bkg=bkg_sky, binning=2,
    n_threads=2)

Let’s take a look at the velocity map. We can play with the colorbar limits with the clims argument. Please note that the flux plot is automatically scaled by log10. However, the velocity and broadening maps are not scaled automatically.

lplt.plot_map(flux_map[:,:,0], 'flux', object_name=object_name, filter_name=filter_name, output_dir=cube_dir, header=cube.header, clims=[-18, -14])

And let’s see what this looks like!

We can also plot the velocity and broadening.

lplt.plot_map(vel_map[:,:,0], 'velocity', object_name=object_name, filter_name=filter_name, output_dir=cube_dir, header=cube.header, clims=[0, 100])

lplt.plot_map(broad_map[:,:,0], 'broadening', object_name=object_name, filter_name=filter_name, output_dir=cube_dir, header=cube.header, clims=[0, 40])

The resulting data maps will be placed in a folder called Luci_outputs. Inside there, you will find additional folders containing the Flux, Amplitude, Velocity, and Broadening maps for each line and their uncertainties (if calculated).