v0.7.2

Changed

• dsigma.stacking.lens_magnification_bias can now be used to compute the bias in the tangential shear
• mean source redshift now takes into account n(z)'s passed to dsigma.precompute.precompute

Fixed

• incompatibility with numpy 1.24
• bug in dsigma.stacking.tangential_shear when random_subtraction=True
• error in tomographic redshift bin assignment for KiDS, sources with photo-z's at the bin edges were assigned to the wrong tomographic bin, this biased KiDS lensing measurements by order 2%