feat: switch to nalgebra

[Smirkey]

Solves #57

[codspeed-hq]

CodSpeed Performance Report

Merging #59 will degrade performances by 32.74%

_{Comparing experimental/nalgebra (4939498) with main (f18079d)}

Summary

❌ 3 regressions
✅ 6 untouched benchmarks

⁉️ 140 dropped benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	main	experimental/nalgebra	Change
⁉️	test_box_convert_cxcywh_xywh[float32]	31.7 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[float64]	36 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[int16]	29.3 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[int32]	31.5 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[int64]	32.7 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[uint16]	28.9 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[uint32]	31.4 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[uint64]	36.1 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xywh[uint8]	28.7 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[float32]	33 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[float64]	36.3 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[int16]	29.3 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[int32]	31.6 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[int64]	33 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[uint16]	28.9 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[uint32]	31.4 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[uint64]	36.2 µs	N/A	N/A
⁉️	test_box_convert_cxcywh_xyxy[uint8]	28.8 µs	N/A	N/A
⁉️	test_box_convert_xywh_cxcywh[float32]	31.7 µs	N/A	N/A
⁉️	test_box_convert_xywh_cxcywh[float64]	32.9 µs	N/A	N/A
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.