Skip to content

Added a all_positivity_from_dataspecs key to get the union of positivity datasets

Emanuele Roberto Nocera requested to merge vp_union_of_positivities into master

Created by: scarlehoff

As discussed in #593 (closed) it would be nice to have a feature enabling the possibility of doing a comparefit where the two fits were done with a different set of positivity datasets. I've done so by creating a produce_all_positivity_from_dataspecs method akin to produce_matched_positivity_from_dataspecs.

This new feature can be used with the key all_positivity_from_dataspecs.

An example report in which this has been used can be found here: https://vp.nnpdf.science/UM6m_h28Thuf18ocrJENjg==/ where each report use a different positivity set.

Note: this is also a textbook example of cargo cult programming as I don't fully comprehend the logic behind produce_matched_positivity_from_dataspecs, in particular this loop:

for ispec, spec in enumerate(dataspecs):            
    #Passing spec by referene                       
    d = ChainMap({                                  
        'posdataset':       all_names[ispec][k],    
                                                    
        },                                          
        spec)                                       
    l.append(d)

Merge request reports