Skip to content

Instantly share code, notes, and snippets.

@tspriggs
Last active March 14, 2018 17:23
Show Gist options
  • Save tspriggs/989b57a495ab2831a021fb376e18fabc to your computer and use it in GitHub Desktop.
Save tspriggs/989b57a495ab2831a021fb376e18fabc to your computer and use it in GitHub Desktop.
custom model to fit a 2D moffat model to data, equate that to F_OIII_xy, then pass it to a 1D custom Gaussian equation to get a model spectrum. The idea is to have x, y and l (wavelength) as inputs, then the amplitude is fixed as it is calculated. When I run it I normally get mismatched dimension errors: wavelength is a 271 long list, and then x…
def Moffat_3d_test(x, mean, stddev, Gauss_bkg, Gauss_grad,
Moffat_amplitude, x_0, y_0, gamma, alpha, Moffat_bkg):
# Moffat
rr_gg = ((x_fit - x_0)**2 + (y_fit - y_0)**2) / gamma**2
F_OIII_xy = Moffat_amplitude * (1 + rr_gg)**(-alpha) + Moffat_bkg
# Prep for Gauss 1D
Gauss_std = np.sqrt(stddev**2 + std_MUSE**2)
A_OIII_xy = F_OIII_xy / (np.sqrt(2*np.pi) * Gauss_std)
check_1.append(A_OIII_xy)
model_spectra = []
for Amp in A_OIII_xy:
model_spectrum = (Gauss_bkg + (Gauss_grad * x) + np.abs(Amp) * np.exp(- 0.5 * (x - mean)** 2 / Gauss_std**2.) +
(np.abs(Amp)/3) * np.exp(- 0.5 * (x - (mean - 47.9399))** 2 / Gauss_std**2.))
model_spectra.append(model_spectrum)
return model_spectra
#%%
new_model = Model(Moffat_3d_test, independant_vars=["x"])
test_data = np.array(PNe_spectra_list[0][202])
pars = new_model.make_params(mean=5007., stddev=0.9, Gauss_bkg=1., Gauss_grad=0.1,
Moffat_amplitude=20., x_0=8., y_0=8., gamma=4.47, alpha=2.39, Moffat_bkg=10)
pars["alpha"].vary = False
pars["gamma"].vary = False
pars["stddev"].vary = False
#pars["stddev"].min = 0.15
#pars["stddev"].max = 1.0
result = new_model.fit(test_data, x=wavelength, params=pars)
best_fit = result.best_fit
test_1 = best_fit[0]
plt.plot(wavelength, test_data, "k")
plt.plot(wavelength, test_1, "r-")
@tspriggs
Copy link
Author

tspriggs commented Dec 1, 2017

coordinates = [(n,m) for n in range(16) for m in range(16)]

x_fit = np.array([item[0] for item in coordinates])
y_fit = np.array([item[1] for item in coordinates])

x= [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,....] etc. up to 16
y=[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,.....] etc. this is done 16 times

l (or wavelength) is a wavelength range of 271 elements long

@tspriggs
Copy link
Author

tspriggs commented Dec 1, 2017

I also need to figure out where the data is input into the model as well, but this should give you an idea of the process early on

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment