When reading Echelle objects from .ech files, PyReduce ignores the "MASK" property and creates a mask from the "COLUMNS" property.
Conversely, when saving to .ech, PyReduce will save the generated mask as well as the column ranges it came from, not checking if they are redundant.
In practice, this means that the file size grows by ~10% the first time .ech files is loaded and saved, which doesn't seem right.
Happy to work on this, but first I would like to establish what is the desired behavior. My take would be this:
- Reading: Apply the mask, if it exists, otherwise generate mask from column ranges as now, or perhaps even apply column ranges to mask.
- Writing: If the Echelle object has both mask and columns defined, check if they correspond and drop the mask if they do.
When reading
Echelleobjects from .ech files, PyReduce ignores the "MASK" property and creates a mask from the "COLUMNS" property.Conversely, when saving to .ech, PyReduce will save the generated mask as well as the column ranges it came from, not checking if they are redundant.
In practice, this means that the file size grows by ~10% the first time .ech files is loaded and saved, which doesn't seem right.
Happy to work on this, but first I would like to establish what is the desired behavior. My take would be this: