excel gives wrong trend?!

[MRA]

Hi guys, been a while

So I have a set of data here, and I need to generate a best fit polynomial based on it.
So I graph the data and tell excel to give me a polynomial based on it. Turns out that I needed a fifth order polynomial to get the R squared to something I was comfortable with.
The problem is that when I try to test it, that is, write the equation it gives into the spreadsheet and plug in x-values to compare it to what it shows on the graph, the rests I get are WILDLY divergent.

Here's my data

X Y
0.597 295.186
0.601 315.115
0.604 334.512
0.608 336.687
0.611 332.346
0.615 335.401
0.618 371.120
0.622 373.132
0.625 377.151
0.628 372.621
0.632 406.189
0.635 483.511
0.639 519.613
0.642 515.667
0.646 574.437
0.649 596.661
0.653 556.213
0.656 519.006
0.660 483.511
0.663 447.849
0.667 407.783
0.670 369.045
0.674 331.912
0.677 298.255

The eqn is gives is

y = 5E+09x5 - 2E+10x4 + 2E+10x3 - 1E+10x2 + 4E+09x - 6E+08

which as you'll see, generates values that are just goofy, and in most cases 6 orders of magnitude too great (300000000 as opposed to 300).

I've tried entering the formula in manually and copy and pasting it from the chart a number of times and I still get the same result. Anybody know what's goin on?

Thanks a lot guys, you've always come through in the past

 

[Andrew Poulsom]

Right click the equation on the chart and choose Format Data Labels. On the Number tab choose Number with 5 decimal places. You should now get:

5225684650.04102x5 - 16626214733.34560x4 + 21145197930.36260x3 - 13437287685.30730x2 + 4266745083.69304x - 541577724.53373

If your x value is in A2, the formula:

=5225684650.04102*A2^5-16626214733.3456*A2^4+21145197930.3626*A2^3-13437287685.3073*A2^2+4266745083.69304*A2-541577724.53373

will return 286.38. Copy the formula down for the remaining y values.

 

[MRA]

Wow. Just Wow.

You're awesome man thanks a bunch