perf: speed up SVD (~4.7x) and symmetric EVD (~1.7x) via transposed internal storage#167

lpatiny merged 10 commits into

mainmljs/matrix:mainfrom

speedmljs/matrix:speedCopy head branch name to clipboard