How to get a polinom trend eq. for data series not starting at 0?
 

I'm trying to fit a 3rd degree trend line to a data series with independent
values starting from -8 to 10. The 3rd degree equation that I get keeps
assuming that the first independent data point is zero, and so, the
y-intercept of the equation is not the value for -8. These are the data
points,

m Freq (cm-1)
-8 2107
-7 2112
-6 2116
-5 2120
-4 2124
-3 2128
-2 2132
-1 2136
0 2139
1 2147
2 2151
3 2155
4 2158
5 2162
6 2166
7 2169
8 2173
9 2176
10 2180
And so, the y-intercept at x=0 should be close to 2139 instead of the 2104
value that I get. How to make excel recognize that the first data point is
not zero?
Please help,
Juan

I get y=-0.0041x^3-0.0121x^2+4.3565x+2141.2 for the trendline equation
With LINEST (see my site www.stfx.ca/people/bliengme/ExcelTips) I get the
same:
-0.00412 -0.01211 4.356495 2141.183


What makes you think the 'real' value of the intercept is 2139?
Please explain more what you mean by "The 3rd degree equation that I get
keeps
assuming that the first independent data point is zero"

By the way, you are working with a "polynomial"

best wishes
--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email

"juannrobless" wrote in message
...
I'm trying to fit a 3rd degree trend line to a data series with
independent
values starting from -8 to 10. The 3rd degree equation that I get keeps
assuming that the first independent data point is zero, and so, the
y-intercept of the equation is not the value for -8. These are the data
points,

m Freq (cm-1)
-8 2107
-7 2112
-6 2116
-5 2120
-4 2124
-3 2128
-2 2132
-1 2136
0 2139
1 2147
2 2151
3 2155
4 2158
5 2162
6 2166
7 2169
8 2173
9 2176
10 2180
And so, the y-intercept at x=0 should be close to 2139 instead of the 2104
value that I get. How to make excel recognize that the first data point
is
not zero?
Please help,
Juan

Juan -

You are using the Line chart type (which uses values 1,2,3,... for fitting a
trendline).

Instead, use the XY (Scatter) chart type (which will use -8,-7,...).

- Mike
www.mikemiddleton.com


"juannrobless" wrote in message
...
I'm trying to fit a 3rd degree trend line to a data series with
independent
values starting from -8 to 10. The 3rd degree equation that I get keeps
assuming that the first independent data point is zero, and so, the
y-intercept of the equation is not the value for -8. These are the data
points,

m Freq (cm-1)
-8 2107
-7 2112
-6 2116
-5 2120
-4 2124
-3 2128
-2 2132
-1 2136
0 2139
1 2147
2 2151
3 2155
4 2158
5 2162
6 2166
7 2169
8 2173
9 2176
10 2180
And so, the y-intercept at x=0 should be close to 2139 instead of the 2104
value that I get. How to make excel recognize that the first data point
is
not zero?
Please help,
Juan

Use a XY Scatter chart rather than a line chart. Though they look
similar, only the former supports numerical values on the x-axis.

From just plotting the data, it would appear that a linear fit would be
the most appropriate. You should use LINEST with the Stats option to
see if your results are significant. See Trendline coefficients
(http://www.tushar-mehta.com/excel/ti...efficients.htm).

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Multi-disciplinary business expertise
+ Technology skills
= Optimal solution to your business problem
Recipient Microsoft MVP award 2000-2005

In article ,
says...
I'm trying to fit a 3rd degree trend line to a data series with independent
values starting from -8 to 10. The 3rd degree equation that I get keeps
assuming that the first independent data point is zero, and so, the
y-intercept of the equation is not the value for -8. These are the data
points,

m Freq (cm-1)
-8 2107
-7 2112
-6 2116
-5 2120
-4 2124
-3 2128
-2 2132
-1 2136
0 2139
1 2147
2 2151
3 2155
4 2158
5 2162
6 2166
7 2169
8 2173
9 2176
10 2180
And so, the y-intercept at x=0 should be close to 2139 instead of the 2104
value that I get. How to make excel recognize that the first data point is
not zero?
Please help,
Juan