* Add snapshot selection option to evaluate_network while preserving existing functionality.
- Replace snapindex with more explicit snapshot_name. snapshotindex from config is used to identify the snapshot name instead.
- Refactor logic for snapshotindex allowed values. Add limit to negative index values.

* Update evaluate.py
- Replace snapshots_to_evaluate numpy indexing with list loop
- Raise error if no snapshots are found

* Add snapshots_to_evaluate docstring

* Remove unused distance function

* Update evaluate.py
* Improve snapshots_to_evaluate docstring
* Get training_iterations from snapshot_name
* Rename trainingsiterations -> training_iterations

* Improve snapshot_to_evaluate typehint and docstring.

* Refactor evaluate.py
* Create get_available_requested_snapshots for obtaining specified snapshot names (new feature).
* Create get_snapshots_by_index to handle the original implementation.
* Minor improvements and fixes

* Add tests for new snapshot functions in evaluate.py

* Update evaluate_multianimal.py
* Add snapshots_to_evaluate parameter to multi-animal evaluation
* Minor renaming

* Update python-package.yml

* add evaluate_by_snapshot branch for testing purposes

* Add snapshots_to_evaluate usage in testscript.py

* Update get_snapshots_by_index function
* Better typehint for int or str
* Raise IndexError instead of ValueError

* Add get_snapshots_by_index tests to test_evaluate.py

* Update return_evaluate_network_data function with get_snapshots_by_index usage

* Fix potential reference before assignment by raising ValueError

* Refactor getting existing snapshot file names as sorted list.
* Add list_sorted_existing_snapshots to auxiliaryfunctions.py.
* Update several files to use the newly refactored function.
* Preserve same exception handling, albeit with modified message to be more general and concise.

* Refactor evaluate_multianimal.py
* Add new snapshot auxiliary function usage.
* Preserve continuation of loop if no snapshots are found.

* Remove indentation level.

* Rename list_sorted_exisiting_snapshots -> get_snapshots_in_folder
* Update name for all usages
* Minor change to function docstring
* Remove redundant sorting in evaluate_multianimal.py

* Port numpy array to list (not performance critical)

* Add get_snapshots_from_folder tests
* Test snapshots present and returned in order
* Test snapshots not found

* Fix deprecated function names

* Improve snapshots_to_evaluate test in testscript.py

* Correct spelling

* Add per_keypoint_evaluation to evaluate_multianimal_full function call (woops!)

* minor linting

---------

Co-authored-by: Mackenzie Mathis <[email protected]>
Co-authored-by: Niels Poulsen <[email protected]>