Disaggregate shiftlocus

[tiemvanderdeure]

closes #913

[rafaqz]

The double locus shift is a bit worrying, easy to introduce fp errors? Can we not explicitly create start/center/end lookups instead?

[tiemvanderdeure]

yeah it's not ideal - let's find a better solution

[codecov]

Codecov Report

Attention: Patch coverage is 87.50000% with 1 line in your changes missing coverage. Please review.

Project coverage is 83.25%. Comparing base (a15ebb1) to head (e0752c6).
Report is 166 commits behind head on main.

Files with missing lines	Patch %	Lines
src/methods/aggregate.jl	87.50%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #914      +/-   ##
==========================================
+ Coverage   82.32%   83.25%   +0.92%     
==========================================
  Files          60       58       -2     
  Lines        4357     5387    +1030     
==========================================
+ Hits         3587     4485     +898     
- Misses        770      902     +132     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[tiemvanderdeure]

Okay I gave it a go now with a more proper fix and added some tests to show what my goal is here.

Some of the tests fail and I'm not totally sure how to fix them yet, will look more at it later.

But it also seems the tests 1) still use some old (deprecated?) syntax (e.g. passing a locus to disaggregate), and 2) mostly test point Rasters, where I think most of the actual usage of (dis)aggregate will be on interval sampling. It's also much more intuitive what disaggregation of intervals should look like vs disaggregation of points.