ds9norm

This file implements a matplotlib Normalize object which mimics the functionality of image scaling functions in ds9. It provides several ways of configuring image scaling:

Clipping the low and high intensities at specified percentiles (clip_lo, clip_hi)
Specifying an intensity transfer function for ramping from black to white (stretch)
Setting the mid-point of this transfer function (bias, akin to dragging the right mouse button left/right in ds9)
Setting how sharp the transition from black-white is (contrast, akin to dragging the right mouse button up/down in ds9)

Examples

See the example notebook

data = fits.getdata('M51.fits')
figure, axs = plt.subplots(ncols=4, nrows=4, squeeze=False, tight_layout=True)

for ax, stretch in zip(axs[0], ['linear', 'sqrt', 'arcsinh', 'log']):
    ax.imshow(data, norm=DS9Normalize(stretch=stretch))
    ax.set_title(stretch)

for ax, contrast in zip(axs[1], [0.5, 1, 2, -1]):
    ax.imshow(data, norm=DS9Normalize(contrast=contrast))
    ax.set_title('Contrast = %0.1f' % contrast)

for ax, bias in zip(axs[2], [.2, .5, .8, .9]):
    ax.imshow(data, norm=DS9Normalize(bias=bias))
    ax.set_title('Bias = %0.1f' % bias)

for ax, (lo, hi) in zip(axs[3], [(0, 100), (1, 99), (5, 95), (10, 90)]):
    im = ax.imshow(data, norm=DS9Normalize(clip_lo=lo, clip_hi=hi))
    ax.set_title('%i-%i%%' % (lo, hi))