How to do Response Surface Regression Modelling?
Article ID: KB0073289
Updated On:
| Products | Versions |
|---|---|
| Spotfire Statistica | 13.0 and higher |
Description
How to do Response Surface Regression Modelling in Statistica?
Issue/Introduction
How to do Response Surface Regression Modelling?
Environment
Windows
Resolution
1. For illustration, open Statistica example dataset composit.sta
2. Go to Statistica menu "Statistics | Advanced Models | General Linear", choose "Response Surface regression" as analysis type and "Quick Specs" as specification method. Click Ok
3. In the Response Surface regression analysis dialog, click on "Variable" to specify the Dependent variable, Regressors and Blocking Variable.
4. Continue click on "Options" tab, tick "Lack of fit" checkbox
5. Click Ok to run the analysis.
6. In the Response Surface Regression analysis result dialog, under "Quick" tab, click on "Whole Model R".
The output table "Test of Lack of fit" gives the result of Lack of fit information:
The output table "Test of SS Whole Model vs. SS Pure Error" gives the variance result of the Whole Model and the Pure Error.
The output table "Test of SS Whole Model vs. SS Residual" gives the result of the model R, R-squared, Adjusted R-squared.
7. In the result dialog, under "Summary" tab, click on "Coefficients" to display the Parameter estimates for the model.
8. In the result dialog, under "Report" tab, click on "Pred. equation" to display the prediction equation for the analysis model
9. In the result dialog, under "Profiler" tab, click on the "1" "2" graph buttons to display the 3D surface plot and 2D response plot for this analysis.
User can set the desirability values and curvature for the surface plots.
Click on the question mark "?" on the top right corner of the result dialog to reach the Statistica e-manual for detailed explanations of the parameters in the dialog.
2. Go to Statistica menu "Statistics | Advanced Models | General Linear", choose "Response Surface regression" as analysis type and "Quick Specs" as specification method. Click Ok
3. In the Response Surface regression analysis dialog, click on "Variable" to specify the Dependent variable, Regressors and Blocking Variable.
4. Continue click on "Options" tab, tick "Lack of fit" checkbox
5. Click Ok to run the analysis.
6. In the Response Surface Regression analysis result dialog, under "Quick" tab, click on "Whole Model R".
The output table "Test of Lack of fit" gives the result of Lack of fit information:
The output table "Test of SS Whole Model vs. SS Pure Error" gives the variance result of the Whole Model and the Pure Error.
The output table "Test of SS Whole Model vs. SS Residual" gives the result of the model R, R-squared, Adjusted R-squared.
7. In the result dialog, under "Summary" tab, click on "Coefficients" to display the Parameter estimates for the model.
8. In the result dialog, under "Report" tab, click on "Pred. equation" to display the prediction equation for the analysis model
9. In the result dialog, under "Profiler" tab, click on the "1" "2" graph buttons to display the 3D surface plot and 2D response plot for this analysis.
User can set the desirability values and curvature for the surface plots.
Click on the question mark "?" on the top right corner of the result dialog to reach the Statistica e-manual for detailed explanations of the parameters in the dialog.
Feedback
Yes
No
Powered by
