Closed corentinravoux closed 1 year ago
Is it possible to push another minor/fix version for this fix ?
Yes I think the np.nan orginate from empty SNR bins, and the k, z empty bins were already dealt with a warning not accounted for by the previous PR.
Fix for the previous PR. In certain cases (no chunks in a specific wavenumber and redshift bin), the new changes crashed. Improvement of the masking of np.nan by the way.