def _get_log_like_plus_log_factorial(self, obs, bkg, model, bin_id):
arrs = obs, bkg, model
if self._has_top_hats():
arrs = (np.array((self._top_hat_sum(arr, bin_id),)) for arr in arrs)
return log_likelihood(*arrs)
def _compute_likelihood_biases(self):
for bin_label in self._maptree:
data_analysis_bin = self._maptree[bin_label]
this_log_factorial = np.sum(logfactorial(data_analysis_bin.observation_map.as_partial()))
self._log_factorials[bin_label] = this_log_factorial
# As bias we use the likelihood value for the saturated model
obs = data_analysis_bin.observation_map.as_partial()
bkg = data_analysis_bin.background_map.as_partial()
sat_model = np.clip(obs - bkg, 1e-50, None).astype(np.float64)
self._saturated_model_like_per_maptree[bin_label] = log_likelihood(obs, bkg, sat_model) - this_log_factorial
def get_log_like(self):
"""
Return the value of the log-likelihood with the current values for the
parameters
"""
n_point_sources = self._likelihood_model.get_number_of_point_sources()
n_ext_sources = self._likelihood_model.get_number_of_extended_sources()
# Make sure that no source has been added since we filled the cache
assert n_point_sources == self._convolved_point_sources.n_sources_in_cache and \
n_ext_sources == self._convolved_ext_sources.n_sources_in_cache, \
"The number of sources has changed. Please re-assign the model to the plugin."
# This will hold the total log-likelihood
total_log_like = 0
for bin_id in self._active_planes:
data_analysis_bin = self._maptree[bin_id]
this_model_map_hpx = self._get_expectation(data_analysis_bin,
bin_id, n_point_sources,
n_ext_sources)
# Now compare with observation
bkg_renorm = self._nuisance_parameters.values()[0].value
obs = data_analysis_bin.observation_map.as_partial(
) # type: np.array
bkg = data_analysis_bin.background_map.as_partial(
) * bkg_renorm # type: np.array
this_pseudo_log_like = log_likelihood(obs, bkg, this_model_map_hpx)
total_log_like += this_pseudo_log_like - self._log_factorials[bin_id] \
- self._saturated_model_like_per_maptree[bin_id]
return total_log_like
