Tess-sip

tfetherolf has opened a PR to add in functionality to run on LightCurveFiles! This capability has now been merged in with new tests.

lightkurve` had a bug described [here](https://github.com/lightkurve/lightkurve/pull/982), which has (since been fixed in version 2.0.4)[https://github.com/lightkurve/lightkurve/releases/tag/v2.0.4]. This update fixes TESS-SIP to be compatible with the current version of `lightkurve`, with thanks to nespinoza and tfetherolf for identifying and fixing this bug!

Because this tool is also applicable to finding long periods in K2, it's nice if it works out of the box for K2 targets. I've added the following changes:

* Tool can now accept Kepler/K2 target pixel files, as well as TESS
* Optionally, the tool can also detrend the data using the SFF technique. This is particularly helpful for the motion noise in K2. Use the new `sff` and `sff_kwargs` keywords to run a SIP with SFF detrending.

Below is a demonstration for an M67 target from K2. Note, in this case there is no significant period, because there is the same period identified in the background pixels. However, the motion noise is nicely detrended.


Some users (Jen van Saders and Max Guenther) pointed out a few use cases for TESS-SIP, and I've updated based on their experiences.

Changes:
* TESSCut TPFs can now be run in TESS-SIP: we can now use FFI TPFs in TESS-SIP
* Running background periodograms: It's sometimes difficult to tell apart the true period from TESS periodicities. To help with this, I've added the ability to automatically run the periodogram on the background pixels too, i.e. the pixels outside of the aperture. This means that we have a periodogram of the background at the same time. Any period that is consistent with the background is probably spurious.
* Added the ability to select the number of PCA components.

This release changes the demo target, and ensures that the tool runs from the terminal.

This release converts our demo TESS-SIP tool into a pip installable package.