TMPLS - Quick Tab

Select the Quick tab of the TMPLS dialog box to access options to select variables for the analysis and identify codes for categorical variables (batch, time, and factor).

See also, MSPC Technical Notes and PCA and PLS Technical Notes. For an example of how the ETL options are used in MSPC, see MSPC Raw Data Example.

Element Name Description
Variables Click the Variables button to display a five-list variable selection dialog box, from which you select a single batch, a single time variable, at least one dependent variable, and a minimum of two continuous and/or categorical variables. The batch variable associates an observation to a batch and the time variable identifies the time slot in which the observation was made. In TMPLS analysis, the dependent variable is a quality variable that is measured once in the entire duration of the batch. This measurement is then taken to quantify the quality of the batch product. Note that in batch processes you can have more than one dependent (quality) variable measuring various quality aspects of the end product. The continuous and categorical variables are the independent variable of the TMPLS model. In other words they are the process variables.
Note: in TMPLS analysis, the dependent variables must be selected manually. In this case, the dependent variables are quality variables which can be used to judge whether the end product meets the required specifications and standards. This is in contrast to BMPLS models where the time variable is used as the dependent variable. See PCA and PLS Technical Notes for more details.
Categorical responses (categorical dependent variables) Select this check box to designate the dependent variables as categorical, i.e., to perform a classification analysis. When this check box is selected, the Response codes button (see the description of this option below) will become active, provided you have specified a dependent variable. Note that if this check box is not selected, the response variables will be treated as continuous and, accordingly, the problem will be treated as regression.
Time code for dependent variables Use this option to specify a time slot (per batch) for the dependent variable which will be used by the TMPLS model as the measurement of the quality variable for a batch. For example, if 10 measurements were taken during a single batch run, then selecting 5 will identify the value of the dependent variable taken during time slot 5 as the measure of product quality. This value will then be used for the whole batch. Thus every batch will have a single quality measure (per dependent variable). Quality variables are often hard to measure or simply cannot be measured. When possible, however, these variables are measured a lot less frequently than the process variables.
Batch codes This button is available only if a batch variable has been specified. Click the Batch codes button to display the Select batch codes dialog box, which is used to select codes for the batch variable. The list of codes can be displayed by clicking the Zoom button in this dialog box. Note that you must select at least two codes for the selected variable.
Time codes This button is available only if a time variable has been specified. Click the Time codes button to display the Select response codes dialog box, which is used to select codes for the dependent variables. The list of codes can be displayed by clicking the Zoom button in this dialog box. Note that you must select at least two codes for the selected dependent variable.
Factor codes This option is available after variables have been selected and only if they include at least one categorical variable. Click this button to display the Select factor codes dialog box, which is used to select codes for the categorical factors. The list of codes can be displayed by clicking on the Zoom button in this dialog box. Note that you must select at least two codes for the selected categorical factor.
Response codes This button is available only if a dependent variable has been specified and the Categorical responses check box (see above) is selected. In this case, the dependent variable is assumed to be categorical in nature, and the objective of the analysis is then to correctly classify the cases (observations) into groups specified by the dependent variable. Click the Response codes button to display the Select response codes dialog box, which is used to select codes for the dependent variables. The list of codes can be displayed by clicking the Zoom button in this dialog box. Note that you must select at least two codes for the selected dependent variable.
Data in raw form When this check box is selected, MSPC will utilize Extract, Transform, and Load (ETL) functionality to convert "raw" datetime values (e.g., 3/17/2009 5:32 PM) into discrete time interval steps (e.g., 1, 2, 3). When the check box is cleared, MSPC will assume the data set already contains datetime values in the form of discrete time interval steps (e.g., 1, 2, 3).
Time intervals This option is enabled when the Data in raw form check box is selected. The default value for (N-equal) time intervals is 50, with a lower bound of 2 and an upper bound of 100,000. Intervals are computed using the method specified by the Computation of percentiles option, located on the Limits tab of the Options dialog box [accessed by selecting Options from the Tools menu (classic menus) or by clicking Options in the Tools group of the Home tab (ribbon bar)]. The default method is Empirical Distribution Function w/Average.