I'm reading that in Excel you can't use the regression formulas directly but must use the Solver to constrain your regression coefficients.
This may be helpful.
Hi,
I am a postgraduate student carrying out air quality research for my thesis. I have determined a best fit quadratic equation with a zero intercept (y = 1.1133x^{2} – 0.0844x) using OLS regression analysis (MS Excel 2013) for a small dataset. As you can see, this has resulted in a negative linear coefficient which raises some difficulties with physical constants that cannot be less than zero. I need to rerun the regression fit with the constraint that both coefficients have to be ≥0 (i.e. positive), retaining the zero intercept, and have been advised that this is possible using MS Excel Solver. Sadly I cannot see how to achieve this. The only other statistical packages I have access to are Minitab (v16) or the openair package in R. Any ideas?
PS Apologies if I have posted to the wrong forum - I am a complete rookie at this game and more than happy to accept redirection.
Many thanks