Object-oriented data analysis

I'm working on converting my historically functional style of working with images and datacubes to an object-oriented programming (OOP) style.

UNDER CONSTRUCTION The conversion isn't yet complete and still has a few redundancies

The new object-oriented approach provides:

• Better code organization and maintainability
• Encapsulation of data and methods
• Automatic state management and caching
• Easier extensibility and reusability

New Object-Oriented Classes

DataCube Class (datacube.py)

The main class for handling astronomical datacubes.

from datacube import DataCube

# Load from FITS file
datacube = DataCube(filepath='my_datacube.fits')

# Or create from data arrays
datacube = DataCube(data=data_array, header=fits_header)

# Access properties (automatically computed)
wavelengths = datacube.wavelength_axis
frequencies = datacube.frequency_axis
pixel_xx = datacube.pixel_x
pixel_RA = datacube.pixel_ra

shape = datacube.shape

# Extract subcubes
subcube = datacube.extract_subcube([[100, 200], [50, 100], [50, 100]])


#Get VO images
datacube.get_sdss_image()

Image Class (datacube.py)

The main class for handling astronomical images.

from datacube import Image

# Load from FITS file
image = Image(filepath='my_image.fits')

# Or create from data arrays
image = Image(data=data_array, header=fits_header)

# Access properties (automatically computed)
pixel_xx = image.pixel_x
pixel_RA = image.pixel_ra

shape = image.shape

#Get VO images
image.get_sdss_image()

Running the Demo

Execute the demonstration script to see the conversion in action:

python datacube_demo.py

Dependencies

• astropy - FITS file handling, WCS transformations, units
• numpy - Numerical computations
• scipy - Statistical functions, optimization
• matplotlib - Plotting and visualization

Future Extensions

Things I'm working on adding:

• Simple emssion line identification: For continuum-subtracted cubes