Mismatch between bias and science frames

[cristobal-sifon]

It may just be a matter of finding the correct biases among all that I downloaded. This issue should be kept open. Everything that I might find and list here should be accounted for when implementing #11.

I already found that among all biases there are different binning schemes, encoded by the keyword CCDSUM.

The current error is

ERROR - GIREDUCE: image gS20170917S0108 has not been overscan subtracted.   
                  However, BIAS image gS20170914S0179_bias appears          
                  to have been created from biases that were                
                  overscan subtracted.