2106 RIX Lars : Some DRL Functions with unclear description of Outputs

[astronomyk]

https://jira.eso.org/browse/MET-2106

Section 7.3 describes the interface of a range of DRL functions with quite a level of detail.
For some of these functions, the 'Outputs' is not clear, e.g.
7.3.1: "Detrended images [px]" is mostly a copy of the general description (and the input is just a single image).
7.3.2: "Building sky background from images [px]" - same.
7.3.3: "Removing sky background from images [px]" - same.
Other DRL functions list the planned PRO.CATG, which is quite a bit more helpful.

Also, 7.3.6 has "Updated flat field mask" which suggests an in-place operation, although the input is described as const. Same for 7.3.7.

To start the development with unambiguous requirements, the Outputs of the DRL functions should be clarified and checked for inconsistencies. With the General Description already in place, the PRO.CATG (or in its absence simply e.g. hdrl_image * output) would be helpful here.

[hugobuddel]

Yes section 7 is very messy... Almost all of them are wrong. But I believe it would be better to first write at least one recipe and then develop a standardized way to define these functions.

Nevertheless, we could have done a better job. But well, we didn't have time for that.

Proposed reply (no need to comment on it; I'll let it rest a bit so I can answer tomorrow):

Yes you are right; many DRL functions aren't described well enough to be unambiguous. We will go through all DRL functions and ensure their consistency, and properly list their input and output.

[hugobuddel]

Answered.

[JenniferKarr]

I did a first past of standardizing/updating as follows

all input/output have a data type attached to them, ie

• hdrl_image
• hdrl_catalogue_result
• hdrl_imagelist
• hdrl_spectrum1D
• cpl_table

Outputs have either a short description, or the name of the corresponding FITS file as given in Chapter 8. Note that the functions may not write the FITS file themselves. In general, a function that produces a single definite output (like a flat) has a name, ones that are more general (like subtracting flats) have a description. All routines now have an output listed.

Removed the "const" declaration from in-place functions (like subtracting the flat field). The updated image, etc. is listed separately in output, with a description.

Gave more informative variable names for input/output for non-generic input/output.

Right now, the fields used are Input (main input), Other Input (paramters), and Output. Other INput probably needs a second pass to move calibration files to that category.

Fixed some spacing issues.

Restored the LSS section heading, which had been commented out. This will change some section numbers.

[JenniferKarr]

Changes are in branch MET2106

[JenniferKarr]

Final changes for the RIX; added PRO.CATG to outputs in section 7.3 (as referenced in the RIX), with a bit of tidying.