I have a small 4-pair (x,y) data set that I'm trying to extrapolate using a 2nd order curve-fit. I have the desired value of "y" and I want the corresponding "x", which may be a negative number. If I plot the data in an Excel chart, and "forecast" the 2nd order trendline, I get a reasonable answer. If I use the LINEST function, calculate the polynomial coefficients and solve the quadratic equation, I get a reasonable answer if I control the (+/-) option on the quantity +/-SQRT(B^2-4AC). There's two real solutions (roots) to the quadratic. Using the -SQRT(B^2-4AC) option, yields a result which is compatible with the trendline forecast on the Excel chart.
However, when I use the TREND function to extrapolate the dataset as a 2nd order polynomial, the result is limited to the value which is produced when solving the quadratic equation with the +SQRT(B^2-4AC) formulation. Is there anyway for me to control the TREND function to avoid this ambiguity?