run_NS_NestedSampleIt_Builder_dimerization_model.py

'''
Generated by nestedsample_it
Gleipnir NS run script for dimerization_model.py
'''
from dimerization_model_builder import builder
import numpy as np

from gleipnir.pysb_utilities import NestedSampleIt

if __name__ == '__main__':
    # Initialize PySB solver object for running simulations.
    # Simulation timespan should match experimental data.
    tspan = np.linspace(0,1, num=51)

    # USER must add commands to import/load any experimental
    # data for use in the likelihood function!
    experiments_avg = np.load('dimerization_model_dimer_data.npy')
    experiments_sd = np.load('dimerization_model_dimer_sd.npy')

    # Setup the Nested Sampling run
    population_size = 100
    observable_data = dict()
    time_idxs = list(range(len(tspan)))
    observable_data['A_dimer'] = tuple((experiments_avg, experiments_sd, time_idxs))
    # Initialize the NestedSampleIt instance with the model details.
    sample_it = NestedSampleIt(builder.model, observable_data, tspan, builder=builder)
    # Now build the NestedSampling object. -- All inputs are
    # optional keyword arguments.
    nested_sampler = sample_it(ns_version='built-in',
                               ns_population_size=population_size,
                               ns_kwargs=dict(),
                               log_likelihood_type='snlpdf')
    print(len(nested_sampler.sampled_parameters))
    # Then you can run the nested sampler.
    log_evidence, log_evidence_error = nested_sampler.run(verbose=True)
    # Retrieve the evidence
    evidence = nested_sampler.evidence
    print("evidence: ",evidence)
    print("log_evidence: ", np.log(evidence))