When removing the code that transposed WCS axes (which was a kludge for some OCP data, but not all) it reintroduced an error caused by hardcoding the dataset axes in the headers. To alleviate that, this PR dynamically sets the dataset headers based on the L0 headers so that dataset inventory creation doesn’t freak out downstream.
each axis type is checked to ensure it is the same for all frames in the dataset
each axis type is saved as a constant
when writing the L1 files, these constants are used to determine which dataset axis should be which
When removing the code that transposed WCS axes (which was a kludge for some OCP data, but not all) it reintroduced an error caused by hardcoding the dataset axes in the headers. To alleviate that, this PR dynamically sets the dataset headers based on the L0 headers so that dataset inventory creation doesn’t freak out downstream.
each axis type is checked to ensure it is the same for all frames in the dataset
each axis type is saved as a constant
when writing the L1 files, these constants are used to determine which dataset axis should be which
Â