Hi, zero value in the database for dependance and independance variables are counted as valid sample in the multiple regression analysis.
I can't clean the zero valua in the database, are there any method to adjust the forecasted value based on the actual sample size after removing the zero value?
Before performing multiple regression on the data that you import from the datbase, you could first mask all rows that have the zero value? Rows that are masked will be ignored in the computation. To mask a row, you can just right-click on the row header to the extreme left and select the Mask menu item in the context drop-down menu. Or you can use the Mask tool, which can be turned on (if not already shown) using the View->Toolbars dialog.
Remove Invalid Sample in Multiple Regression
Posted - 11/09/2004 : 11:23:12 PM
