Another stumper ... I am running a multiple regression with two independent variables. The regression results look good, except that one coefficient is positive and the second is negative. I need both coefficients to be positive because I am working with a system where I cannot have negative units (it is acceptable for the intercept constant to be negative). Is it possible to force the coefficients to be positive? Or at least is there some way to examine all combinations of coefficients and select the best pair that is positive?