combine_significance_maps#

gammapy.estimators.utils.combine_significance_maps(maps)[source]#

Computes excess and significance for a set of datasets. The significance computation assumes that the model contains one degree of freedom per valid energy bin in each dataset. The method implemented here is valid under the assumption that the TS in each independent bin follows a Chi2 distribution, then the sum of the TS also follows a Chi2 distribution (with the sum of the degrees of freedom).

See, Zhen (2014): https://www.sciencedirect.com/science/article/abs/pii/S0167947313003204, Lancaster (1961): https://onlinelibrary.wiley.com/doi/10.1111/j.1467-842X.1961.tb00058.x

Parameters:
mapslist of FluxMaps

List of maps with the same geometry.

Returns:
resultsdict

Dictionary with entries:

  • “significance” : joint significance map.

  • “df” : degree of freedom map (one norm per valid bin).

  • “npred_excess” : summed excess map.

  • “estimator_results” : dictionary containing the flux maps computed for each dataset.

See also

get_combined_significance_maps

same method but computing the significance maps from estimators and datasets.