byx gives the expected value of y when x is independent and bxy gives the expected value of x when y is independent, where byx and bxy are the coefficients of linear regression.
Explanation:
r = √{byx×bxy } is the Pearson's coefficient of correlation.