The default tolerance for the inner optimization of semiquantitative observables was set too low. Setting it to the default L-BFGS-B tolerances.
Additionally, there was a small bug in the inner calculator collector: the quantitative_data_mask of the edatas was not taking into account that some entries of edatas can be nan. This led to some failures and nan values in the calculation of the quantitative result.
The default tolerance for the inner optimization of semiquantitative observables was set too low. Setting it to the default L-BFGS-B tolerances.
Additionally, there was a small bug in the inner calculator collector: the
quantitative_data_mask
of the edatas was not taking into account that some entries of edatas can be nan. This led to some failures and nan values in the calculation of the quantitative result.