mmirs_pipeline_taskfile: Handling of calibration datasets with calibration-identification algorithm

[astrochun]

This will be an open issue to report and fix the calibration-identification algorithm.

[astrochun]

For full nights with both longslit and imaging for the same target, the algorithm can be confused if the name has a lot of underscores. e.g., 'target_J' will be identified as 'target' in subsequent process as underscore is often use for obj0: List of strings that combines name, aperture, filter, and disperse. To fix this, I decided to change my separator to a colon:

This has been implemented in:
organize_targets: 33b1d79, 50c242b
get_calib_files: b2365a5
create: 847d28b, 7973399, 2fb6b65
generate_taskfile: c1abcae

[astrochun]

A common issue is with multi-datasets of telluric star. This can happen either before or after science observations. Generally it is because of different exposure times. For those before science, it is OK since code by default uses the ones closest to the first science frame. However, for those after, it uses the first set, which can either have too few counts or be too hot/saturated.

This issue affects a number of observations for UTC night: 2018.0430, 2018.0505, 2018.0508

[astrochun]

Until I improve the algorithm completely, I have implemented an interactive option by setting inter=True.

Added in:
create (580537b)
get_tellurics: (ded0a3f, b63eaa5)
handle_tellurics (ded0a3f, 7760943)

Note that this new feature has not been fully tested out.

[astrochun]

I found a new bug (though it is a because we did not consider this possibility in a nights worth of data). For some nights, it is possible that a telluric star is observed in different modes with the same exposure time. This is not identifiable and can group longslit with MOS data.

I've implemented some changes to fix this issue:
handle_tellurics: 7f75047, dce0dec, b1e89b1, 1896d59, e53b7f6