Trendline error

[RuiPedro]

I have a problem with the trendline tool in excel 2010.

I fit perfectly a 5th order polynomial into my x-y scatter chart but then, when I use the equation supplied by excel for that specific trendline, the values obtained are way off (see attached file).

When I create a new chart with the equation given by excel, it is obviously completely different from the original one. And I need that equation to be correct because I want to use it for a model I'm developing in a FEM software.

As far as I know and have seen, this problem is might be associated with the Excel Tips "Polynomial regression. How can I fit my X, Y data to a polynomial using LINEST?", but I'm having some trouble in using that function.

Why a perfectly matching 5th order polynomial trendline gives an equation with an output as wrong as it is? How to explain such a deviation?

How to obtain a really representative equation using this tool?

Best regards,

Rui Pedro

[RuiPedro]

OK, after some trial and errors I think I have the solution for my problem using LINEST:

=LINEST(E5:E15;D5:D15^{1,2,3,4,5})

I'm not sure, yet, what the numbers inside the {} are but I'm guessing it should be the same as the polynomial order, so, in this case, it's 5.

So the question is, why is trendline giving such an off result? Isn't it about time to fix this?