Giving the isophotal parameters as initial guesses while fitting disk+bulge decompositions

[angelogeorge]

I am unbale to find an option to use the elongation/ellipticity from the isophotal measurements as initial guesses while modelling the galaxy. I am currently using a constant initial parameter for aspect ratio. Just like lambda function for radius (o.radius) or angle (o.angle), is there a similar option for ellipticity/axis ratio?

[mkuemmel]

Its o.aspect_ratio, such as:
disk_ratio = FreeParameter(lambda o: o.aspect_ratio, Range((0.1, 1.0), RangeType.EXPONENTIAL))

It's being used in the examples 01_basics and 02_mband in the test data set stored at:
https://cloud.physik.lmu.de/index.php/s/3K4KemBsw5y9yqd

Hope this helps.

[angelogeorge]

Thanks!

My issue is that I cannot perform bulge+disk decomposition on HSC images as effectively as GALFIT. I performed extensive simulations, all of which show that the results are not that great. I was thinking if I gave the o.aspect_ratio as the initial parameter, it would help, but in reality, the results remained more or less the same.

I am not sure if I should start a new thread for this issue. However, if anyone can give me an insight, that would be helpful.

PS: The results seem to be a little better when I adopt "set_modified_chi_squared_scale(0.5)"

Here is a snippet of what I do:

Sersic Models

group = mesgroup

flux = get_flux_parameter()
mag = DependentParameter(lambda f, zp=mag_zp[band]: -2.5 * np.log10(f) + zp, flux)
BT = FreeParameter(0.5, Range((0.001,1.0), RangeType.EXPONENTIAL))
flux_B = DependentParameter(lambda f,r: fr, flux, BT)
mag_B = DependentParameter(lambda f, zp=mag_zp[band]: -2.5 * np.log10(f) + zp, flux_B)
flux_D = DependentParameter(lambda f,r: f(1-r), flux, BT)
mag_D = DependentParameter(lambda f, zp=mag_zp[band]: -2.5 * np.log10(f) + zp, flux_D)
x_B, y_B = get_pos_parameters()
x_D, y_D = get_pos_parameters()

nsersic_B = FreeParameter(n_prior_ini, Range((n_prior_min, n_prior_max), RangeType.EXPONENTIAL))#.LINEAR))

rad_B = FreeParameter(lambda o: 0.1 * o.radius, Range(lambda v, o: (0.01 * o.radius, 2 * o.radius), RangeType.EXPONENTIAL))
rad_D = FreeParameter(lambda o: o.radius, Range(lambda v, o: (0.1 * o.radius, 10 * o.radius), RangeType.EXPONENTIAL))
rad_B_arc = DependentParameter(lambda r: r0.168, rad_B)
rad_D_arc = DependentParameter(lambda r: r0.168, rad_D)

angle_B = FreeParameter(lambda o: o.angle, Range((-np.pi, np.pi), RangeType.LINEAR))
angle_B_deg = DependentParameter(lambda x: np.fmod(x,np.pi/2.0)/np.pi*180.0, angle_B)
aspect_B = FreeParameter(lambda o: o.aspect_ratio, Range((0.01, 1.0), RangeType.EXPONENTIAL))

angle_D = FreeParameter(lambda o: o.angle, Range((-np.pi, np.pi), RangeType.LINEAR))
angle_D_deg = DependentParameter(lambda x: np.fmod(x,np.pi/2.0)/np.pi*180.0, angle_D)
aspect_D = FreeParameter(lambda o: o.aspect_ratio, Range((0.01, 1.0), RangeType.EXPONENTIAL))

prior on n

s0 = 8
sc = DependentParameter(lambda st: np.exp(st - s0), nsersic_B)
add_prior(sc, 0.0, 1.0)

Configuration file

magnitude-zero-point = 27
detection-minimum-area = 10
background-cell-size = 256
smoothing-box-size = 3
segmentation-filter=default.conv
detection-threshold = 2.0
partition-multithreshold = true
partition-minimum-contrast = 0.005
partition-threshold-count = 64
weight-type=variance

Here are the results for HSC i-band:_