Penguins

**Full Changelog**: https://github.com/yongrenjie/penguins/compare/v0.5.0...v0.5.2

Functionally not that many improvements (see full changelog below for details). This release is mainly done to coincide with the [GENESIS paper](https://github.com/yongrenjie/genesis) so that figures from the paper may be reliably reproduced.

**Full Changelog**: https://github.com/yongrenjie/penguins/compare/v0.4.2...v0.5.0

Very small improvements here.

- You can now access different components of the spectrum using `ds.proc_data(component=X)`, where `X` is `"real"` / `"imag"` (1D) or `"rr"` / `"ri"` / `"ir"`/ `"ii"`(2D).
- You can also get rid of the group delay (technically, circularly shift it to the end of the FID) by using `ds.raw_data(shift_grpdly=True)`. **nb** This may be deprecated in a future version, in favour of a keyword argument which provides multiple different methods of processing the group delay.
- Error messages are slightly more user-friendly.

This is also mainly a 'reproducibility' release for the figures in the revised [nmrpoise](https://github.com/foroozandehgroup/nmrpoise) paper.

**Changes**

- Fixed `penguins.mkinset()` to actually work.
- The `penguins.private` module no longer exists. It shouldn't have been here anyway. (The code was moved to [`yongrenjie/aptenodytes`](https://github.com/yongrenjie/aptenodytes).)

This release is meant as a 'reproducibility snapshot', so that the plotting / processing scripts in my NOAH seHSQC paper can be reliably run. (See [DOI:10.1016/j.jmr.2021.107027](https://doi.org/10.1016/j.jmr.2021.107027) for the paper.)

This release is meant for the [nmrpoise](https://github.com/foroozandehgroup/nmrpoise) paper.

**Changes**

- `move_ylabel()` and `move_xlabel()` have been renamed to `ymove()` and `xmove()` respectively, without any retention of backward compatibility.
- `ymove()` now takes three different "configuration styles", namely "topright", "midright", and "topspin". Please see [the documentation](https://yongrenjie.github.io/penguins/howto/move_axes/) for more info.
- `pg.subplots2d()` now takes `height_ratios` and `width_ratios` as arguments. The figsize is determined from either of these if they are passed: each height_ratio of 1 will correspond to an Axes size of 4x4.