Visualization with DeepLenstronomy

deeplenstronomy has a couple built-in visualization functions to help you inspect your simulated data. To see these functions in action, let's start by simulating a dataset.

import deeplenstronomy.deeplenstronomy as dl
from deeplenstronomy.visualize import view_image, view_image_rgb

dataset = dl.make_dataset('data/demo.yaml', store_sample=True, store_in_memory=False, verbose=True)

Entering main organization loop
Organizing CONFIGURATION_1
Organizing CONFIGURATION_2
Organizing CONFIGURATION_3
Organizing CONFIGURATION_4
Generating images for CONFIGURATION_1
	Progress: 100.0 %  ---  Elapsed Time: 0 H 0 M 7 S
Generating images for CONFIGURATION_2
	Progress: 100.0 %  ---  Elapsed Time: 0 H 0 M 7 S
Generating images for CONFIGURATION_3
	Progress: 100.0 %  ---  Elapsed Time: 0 H 0 M 7 S
Generating images for CONFIGURATION_4
	Progress: 100.0 %  ---  Elapsed Time: 0 H 0 M 52 S

view_image()

The view_image() function is built on top of matplotlib.pyplot.imshow() and takes care of some formatting things for you. It is defined as follows:

help(view_image)

Help on function view_image in module deeplenstronomy.visualize:

view_image(image, stretch_func=<function _no_stretch at 0x7ffd2d6f7b90>, **imshow_kwargs)
    Plot an image.
    
    Args:
        image (array):  a 2-dimensional array of pixel values OR a list-like object of 2-dimensional arrays of pixel values 
        stretch_func (func, optional, default=pass): stretching function to apply to pixel values (e.g. np.log10)
        imshow_kwargs (dict): dictionary of keyword arguments and their values to pass to matplotlib.pyplot.imshow

view_image() can be used to display a single image:

view_image(dataset.CONFIGURATION_4_images[0][0])

view_image() can be used to display an array of images, useful for displaying all the bands simultaneously:

view_image(dataset.CONFIGURATION_4_images[0])

You can apply a stretching function to your images before plotting:

import numpy as np
view_image(dataset.CONFIGURATION_4_images[0], stretch_func=np.log10)

You can pass matplotlib.pyplot.imshow() keyword arguments directly to view_image():

view_image(dataset.CONFIGURATION_4_images[0], stretch_func=np.log10, vmin=1, vmax=2, cmap='inferno')

Note: The stretching function is applied before passing keyword arguments to imshow. Therefore, in this example vmin is really $10^1$ and vmax is really $10^2$ if we are considering the unstreteched pisel values.

view_image_rgb()

The view_image_rgb() function is built on top of matplotlib.pyplot.imshow() and astropy.visualization.make_lupton_rgb(). It is defined as follows:

help(view_image_rgb)

Help on function view_image_rgb in module deeplenstronomy.visualize:

view_image_rgb(images, Q=2.0, stretch=4.0, **imshow_kwargs)
    Merge images into a single RGB image. This function assumes the image array
    is ordered [g, r, i].
    
    Args:
        images (List[np.array]): a list of at least 3 2-dimensional arrays of pixel values corresponding to different photometric bandpasses
        imshow_kwargs (dict): dictionary of keyword arguments and their values to pass to matplotlib.pyplot.imshow

You can utilize this function by passing an array of images, ordered from bluest to reddest. This function struggles with noisy and low signal-to-noise images, but is a great check to make sure colors are appearing as expected.

view_image_rgb(dataset.CONFIGURATION_4_images[0], Q=10, stretch=10)

You will often have to tweak the Q and stretch values when using this function. Again, you can also pass matplotlib.pyplot.imshow() keyword arguments to the function, but keywords like cmap are discouraged to allow the make_lupton_rgb() to control the colors displayed.