wrangle
wrangle copied to clipboard
ideas for improving OLS correlation
There could be option to just use the coef as value as opposed to integer or some other weighted option as well :) and maybe std could be used as confidence of some sort?