Index Linest Formula of Excel In JMP possible?

[#1]

Hi All, 

I would like to know that possible to create the Index + Linest formula using JMP formula?

Example of data table:

Temp	Motor Temp	Exchaust Temp
1	35.6	35.2
2	35.9	34.8
3	34.9	34.2

With the index + Linest formula, i would like to get coef1,coef2 and coef3 results as attached in the excel file example. 

Final result as shown in Result column (-2.7619,196.1429 and -3447.1581). 

	Result	Excel Formula
Coef1	-2.7619	INDEX(LINEST($D$2:$D$4, $C$2:$C$4^{1,2}),1,1)
Coef2	196.1429	INDEX(LINEST($D$2:$D$4, $C$2:$C$4^{1,2}),1,2)
Coef3	-3447.1581	INDEX(LINEST($D$2:$D$4, $C$2:$C$4^{1,2}),1,3)

Hope someone can teach me about this. Thank you and appreciate!

[#2]

I assume that you are new to using JMP and you are thinking about JMP as a spreadsheet application. I am going to show you how I would obtain the parameter estimates.

I entered your data into a JMP data table. JMP works with variables, not cells. So each variable is represented by a data column. Each row represents on observation. Your data looks like this:

table.PNG

JMP provides platforms for analysis and graphing. Two continuous variables are explored and fit using the Bivariate platform, but you do not need to know that. Simply click Analyze in the menu bar and select Fit Y by X. Then select the data column and click the Y or X button to assign that variable to the intended analysis role.

dialog.PNG

Click OK and see the default platform. Bivariate starts with a scatter plot of the data.

plot.PNG

Click the red triangle at the top left of the platform and select Fit Polynomial > 2, quadratic.

menu.PNG

This command produces the regression results, including the coefficients.

fit.PNG

See the JMP Help for Bivariate for more information.

[#3]

Just to point out a couple of things. I believe Mark has the X and Y variables mixed up. Motor Temp should be the X, I believe while Exhaust Temp is the Y. Beyond that you will notice that the parameter estimates will not match Excel because JMP will center the polynomial term of the model. This is a good thing to do because it makes the parameter estimates more stable and reliable. The centering reduces multicollinearity in the data, which is a very good thing to do. The centering can be turned off by using the Fit Special command off the red triangle menu that Mark mentioned.