The Bands Data.jmp data table contains measurements from machinery in the rotogravure printing business. The data set contains 539 records and 38 variables. The response Y is the column Banding? and its values are “BAND” and “NOBAND”. You are interested in understanding what properties are most likely to contribute to the response.
1.
Select Help > Sample Data Library and open Bands Data.jmp.
2.
Select Cols > Modeling Utilities > Screen Predictors.
3.
Select columns grain screened to chrome content as the X columns for predictor screening.
4.
5.
Select Banding? as the Y column to predict.
6.
Ranked Columns for Screening Predictors
The columns are sorted and ranked in order of contribution to the response. Columns with rank 21 through 33 in Ranked Columns for Screening Predictors have a negligible contribution to the response. These columns can be excluded from future analyses as they likely have no impact on the response. The most important columns are identified as solvent pct, ink pct, and press.
Launch the Screen Predictors modeling utility by selecting Cols > Modeling Utilities > Screen Predictors.
Screen Predictors Launch Window for Selecting Xs
Screen Predictors Launch Window for Selecting Xs shows the launch window for selecting the predictor X columns. The columns selected here are those you would like to screen.
Screen Predictors Launch Window for Selecting Y
Screen Predictors Launch Window for Selecting Y shows the launch window for selecting the response Y columns.
After selecting the important predictors in the Predictor Screening Report, you can select Analyze > Fit Model. You can then fit a model with the important predictors and leave the unimportant predictors out. In this fashion, the Screen Predictors utility can shorten the modeling process.