FitThinDisk is a Python program to fit the spectral line cubes with a geometrically thin disk model.

git clone https://github.com/yunan77/FitThinDisk.git

• NumPy
• Astropy
• SciPy

FitThinDisk uses scipy.optimize.least_squares to minimizes the value of χ² between model and data by giving the spectral line cubes. Trust Region Reflective algorithm is selected to perform minimization in order to handle the fitting with bounds.

The input cubes must be three dimensional in FITS format, and the spectral axes should be in velocity instead of frequency.

There are seven parameters can be fit by this program. In the initial file params.ini, users can choose which parameters they want to fit by setting variation and can determine the bounds of the parameters.

• qr, power-law index of the rotation velocity as a function of radius, vr0(r/r0)^qr, r0 is an arbitrary reference radius and vr0 is the rotation velocity at the reference radius
• qd, power-law index of the disk intensity as a function of radius, (r/r0)^qd
• r0, reference radius (arcsec)
• x0, horizontal offset of the disk center position from the center position of the image (pix)
• y0, vertical offset of the disk center position from the center position of the image (pix)
• pixsize, size of a pixel in the data cube (arcsec)
• samplesize, size of a pixel in the model cube (arcsec)
• imgrange, desired fitting range on the plane of the sky by pixel index (left, right, down, up)
• dist, source distance (pc)
• bmaj, major axes width of the beam (arcsec)
• bmin, minor axes width of the beam (arcsec)
• bpa, position angle of the beam major axis (deg)

• incl, inclination angel of the disk (deg)(0 for face-on, 90 for edge-on)
• pa, position angel of the major axes of the disk (deg)
• vr0, roataion velocity at the reference radius projected onto the line of sight (km/s)
• linewidth, FWHM of Gaussian line profile (km/s)
• vlsr, source system velocity (km/s)
• ri, inner radii of the disk (arcsec)
• ro, outer radii of the disk (arcsec)

Parameters are read from an input file params.ini.

In [free], the parameters are designed by four parts, initial/fixed value, variation, min and max.

The variation is used to decide whether the parameter should be fit or not. For True, the parameter will be fit and the first value will be taken as initial guess in fitting process; for False, the parameter will be fixed at the first value.

The min and max are designed as bounds of the parameter during fitting process.

[file]
imgpath = /home/user/
imglist = ["cube1.fits",
           "cube2.fits",
           "cube3.fits"
          ]
chanrange = ["0~10",
             "5~30",
             "12~56"
            ]
[fixed]
pixsize = 0.02
samplesize = 0.004
imgrange = 870,940,855,925 
dist = 3080. 
qr = -0.5 
qd = -1
r0 = 0.1 
bmaj = 0.2
bmin = 0.1
bpa = -50
x0 = 5.72
y0 = -9.14

[free]
vlsr = -15.5     ,True,-30,-10
incl = 25        ,True,0,180
pa = -30         ,True,-90,0
vr0sin = 3       ,False,-inf,inf
linewidth = 5    ,False,0,inf
ri = 0           ,False,0,inf
ro = 5.0         ,False,0,inf

python fitdisk.py

Fitting success!
reduced chi square = 1.56
optimized with 12564 effective pixels
[variables]
vlsr        =  -20  +/- 0.01 (init = -15.5)  (bounds = [-30,-10])
incl        =  50   +/- 0.01 (init = 25)     (bounds = [0,180])
pa          =  -60  +/- 0.01 (init = -30)    (bounds = [-90,0])
vr0         =  3             (fixed)
linewidth   =  5             (fixed)
ri          =  0             (fixed)
ro          =  5.0           (fixed)