AttributeError: 'FittingSpecify' object has no attribute 'imageModel'
Closed this issue · 6 comments
Hello,
I am an undergraduate attempting to work through the GaLight galight_hst_QSO.ipynb tutorial before using it on my own work. In the cell "#Setting the fitting method and run" I am getting the following results and Attribute Error:
Computing the PSO ...
10
20
30
40
50
Max iteration reached! Stopping.
-1.0707842721152985 reduced X^2 of best position
-7247.603345812398 logL
13537.0 effective number of data points
[] lens result
[] source result
[{'amp': 1, 'R_sersic': 0.38666619680009284, 'n_sersic': 4, 'e1': -0.022096159868187745, 'e2': -0.4453877445287039, 'center_x': -0.10595217820276748, 'center_y': 0.12255507200599276}, {'amp': 1, 'R_sersic': 0.5234385199161239, 'n_sersic': 1.1694787949844492, 'e1': -0.05072379898905925, 'e2': -0.4994422659942345, 'center_x': -2.8524320644268926, 'center_y': -1.0525580054914583}, {'amp': 1, 'R_sersic': 0.16020498507891734, 'n_sersic': 5.420891413673527, 'e1': -0.4925077338338324, 'e2': -0.27729442818227623, 'center_x': -3.9510706792239767, 'center_y': -1.6123984023508169}] lens light result
[{'ra_image': array([-0.10595218]), 'dec_image': array([0.12255507]), 'point_amp': array([1.])}] point source result
{} special param result
6.391066312789917 time used for PSO
100%|██████████| 130/130 [01:23<00:00, 1.55it/s]
Computing the MCMC...
Number of walkers = 170
Burn-in iterations: 100
Sampling iterations (in current run): 130
84.4948480129242 time taken for MCMC sampling
90.924 total time taken for the overall fitting (s)
============ CONGRATULATION, YOUR JOB WAS SUCCESSFUL ================
Start transfering the Params to fluxs...
AttributeError Traceback (most recent call last)
Cell In[18], line 17
6 fit_run = FittingProcess(fit_sepc, savename = 'HST_result', fitting_level='norm')
7 #For the fitting_level, you can also put ['norm', 'deep'] for the later ['PSO', 'MCMC'] corresplingly.
8
9 #Setting the fitting approach and Run:
(...)
15 # input template: {'n_burn': 50, 'n_run': 100, 'walkerRatio': 10, 'sigma_scale': .1}
16 # if setting_list = [None, None, None], default values would be given
---> 17 fit_run.run(algorithm_list = ['PSO', 'MCMC'], setting_list = None)
19 # Plot all the fitting results, including:
20 # run_diag() : The convergence of the chains.
21 # model_plot(): The model plot (by lenstronomy)
22 # plot_params_corner(): The mcmc corner for all the chains (MCMC should be peformed)
23 # plot_flux_corner(): The flux corner for all the component (MCMC should be peformed)
24 # plot_final_qso_fit() or plot_final_galaxy_fit(): Plot the overall plot (data, model, data-ps, resudal, 1D profile)
25 fit_run.plot_all()
File ~/.local/lib/python3.9/site-packages/galight/fitting_process.py:172, in FittingProcess.run(self, algorithm_list, fitting_level, setting_list, threadCount, refresh)
170 if len(fitting_specify_class.point_source_list) > 0:
171 for j in range(len(fitting_specify_class.point_source_list)):
--> 172 image_ps_j = fitting_specify_class.imageModel.point_source(kwargs_ps_out, k=j)
173 flux_list_quasar.append(np.sum(image_ps_j))
174 flux_list_galaxy = []
AttributeError: 'FittingSpecify' object has no attribute 'imageModel'
My initial suspicion was a version issue, but I have checked all dependencies based on requirements.txt and after updating (or downgrading) to the suggested requirements and I am still encountering this issue. What should I try as my next troubleshooting steps? Thank you for the help!
Hi, I also find this issue on my side. I think this is happening because the recent updates of galight, let me fix this issue. At this moment, you can run 'PSO' and skip 'MCMC'.
Hi, I decided to play with the lenstronomy version after your message. The requirements.txt file listed on GitHub recommends lenstronomy==1.10.3, which I had installed. I uninstalled it and then reinstalled with version 1.11.2 and it worked.
Very excited to make more progress soon. Thanks!
Hi, I am not sure why the lenstronomy version is causing this problem. But glad to hear it works.
I think the previous issue is still by galight, I have fixed it on my side now. I will make a quick galight release today. I recommend you update lenstronomy (1.11.2) and galight (0.1.13, which will be available by tomorrow) to the latest version just in case.
Great, it seems to have been a fluke because the error has already popped up again. Thanks for your work and I'll update to the new version in the morning. Thanks again.
The 0.1.13 version is just released. You can update to the new version and try it out!
Success! Thank you! I'll let you know if I have any other issues. I really appreciate your prompt support