PCA - Options Tab

Select the Options tab of the PCA dialog box to access the options described here.

Element Name Description
Missing data parameters The following options are in the Missing data parameters group box.
Minimum percentage of valid cases per variable (%) Specify the minimum percentage of valid training cases in each variable (i.e., each data set column) in order for the PCA analysis to proceed.
Minimum percentage of valid variables per case (%) Specify the minimum percentage of valid variables in each case (i.e., row of the data set) in order for the PCA analysis to proceed.
Automatically remove the unqualified variables Select this check box to remove the unqualified variables from the analysis and proceed with the rest of the steps for building the PCA model. A variable is unqualified if the number of missing cases exceeds the user-defined percentage or it has no variance (variability). The percentage can be set using the Minimum percentage of valid cases per variable option. If the Automatically remove the unqualified variables check box is cleared, the analysis will stop should the condition for unqualified variables be encountered. Either way, you will be notified that unqualified variables were found.
Automatically remove the unqualified cases Select this check box to remove unqualified cases from the data set chosen for the analysis. A case is unqualified if the user-defined percentage of variable entries are missing. This percentage can be set using the Minimum percentage of valid variables per case option. If the Automatically remove the unqualified cases check box is cleared, the analysis will stop should the condition for unqualified cases be encountered. Either way, the user will be notified that unqualified cases were found.
Maximum data size, in MB Select this check box and specify in the adjacent field the amount of physical memory that the PCA analysis can use. Note that very large data problems (with a large number of variables and cases) may require significant memory and processing resources. Modify the defaults as needed (depending on your computer resources and analysis needs).