Version 1.3.0

Version 1.3.0#

New features#

Apple ARM binaries#

pip and anaconda binaries for apple M processors (python version = 3.11)

C threads#

Introduction of C threads: Starting from version 1.3.0 the JetBase class and all the derived classes, perform the C computation using threads. This increase the computational speed Each time you create a new Jet object, you will get a log noticing how many C threads have been created. The number of threads is automatically determined according to the number of cores and threads of your CPU. You can revert to single thread, or set a custom number of threads, using the JetBase.set_num_c_threads() and passing the number of threads:

from jetset.jet_model import Jet
my_jet=Jet(electron_distribution='lppl')

#to switch back to no threads
my_jet.set_num_c_threads(1)

#to use 8 C threads
my_jet.set_num_c_threads(8)

You can increase or decrease the number of C threads using the same method. It is not advised to exceed the actual number of threads offered by your CPU, since will not increase the speed performance.

Introduction of leptonic synchrotron polarization#

from jetset.jet_model import Jet
my_jet=Jet(electron_distribution='lppl')

nu_range=np.logspace(10,22,50)

p_nu,nuF_nu=jet.eval_synch_pol(nu_range)
#p_nu stores the fractional polarization for each frequency  in nu_range

Custom Models#

Minimizers#

  • Heat map for correlation matrix: added convenience method to plot correlation matrix

  • Added simplex on top of migrad in minuit

MCMC functionalities#

  • labels setting and bounds have been placed in specific methods

    from jetset.mcmc import McmcSampler
model_minimizer = ModelMinimizer.load_model('model_minimizer_minuit.pkl')
mcmc=McmcSampler(model_minimizer)

labels=['N','B','beam_obj','s','gamma0_log_parab']
model_name='jet_leptonic'
use_labels_dict={model_name:labels}

mcmc.set_labels(use_labels_dict=use_labels_dict)

mcmc.set_bounds(bound=5.0,bound_rel=True)

  • added option to plot model and residuals, for the mcmc best fit (obtained setting the pars to their quantiles=0.5 posterior), instead of the frequentist best-fit model, passing plot_mcmc_best_fit_model=True

    p=mcmc.plot_model(sed_data=sed_data,fit_range=[1E11, 2E27],size=100,quantiles=[0.05,0.95], plot_mcmc_best_fit_model=True)

  • customization of lables for plotting

    mcmc.set_plot_label('N',r'$N$')
mcmc.set_plot_label('B',r'$B$')
mcmc.set_plot_label('beam_obj',r'$\delta$')
mcmc.set_plot_label('s',r'$s$')
mcmc.set_plot_label('gamma0_log_parab',r'$\gamma_0$')

EBL models#

EC emission#

  • Improved EC computation for large angles for anisotropic external fields

  • Added convenience methods for conical jet and EC fields:

    • method JetBase.make_conical_jet() class will set parameters dependencies to have conical jet constraining the blob radius

      Important

      if you use a jet model with R depending (i.e. you used JetBase.make_conical_jet()) to perform a temporal evolution (in the JetTimeEvol class), the dependencies on R will be removed, and to have R dependent on the position across the jet axis, use the parameter beta_exp_R in the JetTimeEvol instead. In the next release a more flexible and direct approach will be provided.

    • method JetBase.set_EC_dependencies() class will set parameters dependencies to have scaling relations between BLR and DT radius and disk luminosity

Improved functionalities#

  • Improved dependent parameters: handling of astropy units has been improved the functional dependency of the parameters

  • Improved serialization: saved models will not break if astropy or numba break their interface in future releases

  • Improved functionality to merge light curves jetset.jet_timedep.merge_lcs()

bug fixing#

Contents