pygraphistry
pygraphistry copied to clipboard
perf tests
Presumably there are some cases where we expect cuCat to outperform dirty_cat -- thoughts on how to make into runnable tests, separate from the functional tests?