Pupil preprocessing sequence problems

[dominikbach]

Summary

Users have reported problems with certain sequences of pupil preprocessing steps.

https://github.com/bachlab/PsPM/discussions/573

Problem:

1. Input channels for pspm_pupil_pp and pspm_pupil_correct are checked in pspm_options. Here, it is not allowed to specify channels of type pupil_pp, or to give a channel number. All PsPM functions should at least allow channel numbers.
2. Using the generic data type pupil does not work for pspm_pupil_pp because left and right pupil channel need to be specified individually.

Questions:

1. Do we need pupil_pp and gaze_pp channel types? Generally, PsPM does not use a specific "pp" channel type for preprocessed or filtered data, with the exception of emg_pp which is integrated data that clearly has a different interpretation from the raw data.
2. Do we need the non-standard backroom function pspm_load_single_chan? In any case it should be converted to a regular PsPM function and tested appropriately.

Proposed solution:

1. For the upcoming release v6.1.1, make sure all pupil preprocessing functions create channels of type pupil rather than pupil_pp.
2. For future release, remove pupil_pp channel type altogether (requires changing pspm_init, pspm_load_data, pspm_load_single_chan.
3. For future releases, refactor pspm_load_single_chan into a regular PsPM function.

[dominikbach]

This has been solved by pull requests #623, #621, #620.