This recipe can be used to correct raw data where observations have been taken with inconsistent values of the OBJECT header. The first OBJECT header value found is set for all subsequent files in the group for which it differs, under the assumption that the pipeline will already have grouped the files appropriately.