Fix Inverse for LUFactorized

[jturner314]

This is a hack to fix a.factorize_into()?.inv()? when a is column-major. The reason why this corrects the result is complicated and is related to the fact that the a field in LUFactorized represents the LU factors of A^T in a very confusing way when the input A to the factorization was row-major. The solve modules need an overhaul to properly handle layouts in a clear way.

[codecov]

Codecov Report

Merging #297 (6766c31) into master (082f01d) will increase coverage by 0.31%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #297      +/-   ##
==========================================
+ Coverage   89.01%   89.33%   +0.31%     
==========================================
  Files          71       71              
  Lines        3578     3610      +32     
==========================================
+ Hits         3185     3225      +40     
+ Misses        393      385       -8     

Impacted Files	Coverage Δ
lax/src/solve.rs	100.00% <100.00%> (ø)
ndarray-linalg/src/solve.rs	78.57% <100.00%> (+5.99%)	⬆️
ndarray-linalg/tests/inv.rs	97.82% <100.00%> (+3.38%)	⬆️
ndarray-linalg/src/lobpcg/svd.rs	86.58% <0.00%> (+1.21%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 082f01d...6766c31. Read the comment docs.