Rapidtide

* (docs) Integrated lots of documentation upgrades from Taylor Salo (tsalo).
* (rapidtide) Changed correlation mask to use brain mask as correlation mask if provided.
* (rapidtide) Lowered the default lagminthresh, lagmaxthresh.
* (rapidtide,happy) Improved creation and deletion of canary files.
* (dlfilter) Simplified to support only newer versions of tensorflow 2.X and fix some deprecation problems.
* (atlastool) Fixed 4D processing.
* (showtc) Allowed display of timecourses with different x axes on the same timeline.
* (package) Defined a new filter band - "lfo_tight" from 0.01-0.1Hz (matches a lot of papers).
* (package) Enhanced filter testing.
* (package) Updated requirements to fix tensorflow installation.
* (package) Merged some dependabot PRs.

* (package) Pypi gets mad when you redeploy with the same tag number...

* (retrolagtcs) Now only use one mask (corrmask).
* (retrolagtcs) Now only load the fmri file header to save memory.

* (retrolagtcs) Made a new program to generate lagged timecourses from a lagmap, masks, and a lagtc generator (for fmripost_rapidtide).
* (rapidtide) Added coefficient of variation output.
* (tidepool) Now works with PyQt6.
* (package) Some import cleanups.

* (package) Updated requirements for matplotlib and pyqtgraph to make sure tidepool works (thanks to Hui Zheng for finding this - resolves https://github.com/bbfrederick/rapidtide/issues/168).
* (package) Dropped Python 3.8 support (not supported by the version of pyqtgraph I need).
* (filter.py) You can now specify padtype to all filters.
* (reference) Added the newest version of the HCP-YA mean parameter maps.

* (rapidtide) Estimate and save information about sLFO signal growth over the course of a scan.
* (rapidtide) Now you can have both include and exclude tmasks.
* (rapidtide) Fixed some bugs in memory profiling.
* (rapidtide) Removed some extraneous code to save memory.