Record in the metal matrices the weighted mean redshift of the input to the remapping matrix (i.e. the true absorber redshift) and not the output of the remapping (the 'incorrect' Lya redshift). Both redshifts (input and output) could be useful, but for the purpose of interpreting correctly the metal biases coming from a fit using the metal matrices, the metal redshifts are obviously the correct ones to use. This is what was done in the previous code, so we are fixing here an inconsistency.
Record in the metal matrices the weighted mean redshift of the input to the remapping matrix (i.e. the true absorber redshift) and not the output of the remapping (the 'incorrect' Lya redshift). Both redshifts (input and output) could be useful, but for the purpose of interpreting correctly the metal biases coming from a fit using the metal matrices, the metal redshifts are obviously the correct ones to use. This is what was done in the previous code, so we are fixing here an inconsistency.
This PR addresses issue #1044 .