'  2' '75ư>"HELPNA\NhNN:l|CONTINUE n((pxEXIT&%p=lHALTމPl|dSTEP 1r4߉lg STOP߉X^F%EFSTOP ,,OFFg^w@߉U T߉W O9gp߉.Od9?g|߉CONDUCT BestLines(1,4,1)gSHOW (FillMatrix())psLFeCFSHOW (ANOVA(MainResult,3,2))E(((((((((({\rtf1\ansi\deff0\deftab720{\fonttbl{\f0\fswiss MS Sans Serif;}{\f1\froman\fcharset2 Symbol;}{\f2\fmodern Courier New;}} {\colortbl\red0\green0\blue0;} \deflang1033\pard\tx1200\tx2400\tx4450\plain\f2\fs20 ; GPSS World Sample File - MULTIWAY.GPS \par *********************************************************************** \par * * \par * Example of Multiway Analysis of Variance * \par * based on existing results. See the * \par * sample model OneWay.gps for details * \par * on running Experiments. It is discussed in * \par * Lesson 19 of the Tutorial Manual. * \par * * \par * Here we concentrate on the features and * \par * requirements of the new multiway ANOVA * \par * Library Procedure. * \par * * \par * Predefined Function Keys in MultiWay.gps: \tab * \par * F11 - SHOW (FillMatrix()) * \par * F12 - SHOW (ANOVA(MainResult,3,2)) * \par * * \par * * \par * * \par * Procedure: * \par * 1. Open the Model Multiway.gps * \par * 2. Press Ctrl+Alt+S to Translate the Model and create * \par * the simulation. This defines both the result matrix * \par * and the Table of Residuals. It also initializes the * \par * result matrix to the UNSPECIFIED state. * \par * 3. Press F11 to load run results into the result matrix * \par * 4. Press F12 to perform the Multiway Analysis * \par * of Variance. * \par * 5. Examine the ANOVA Table and Confidence Intervals * \par * in the Journal Window. * \par * 6. Open Window / Simulation Window / Table Window * \par * / MAINRESULT_RESIDUALS * \par * Look for irregularities (i.e. non-Normality) in the * \par * Table of Residuals. * \par * * \par * * \par * The ANOVA Invocation requires 3 arguments. * \par * 1. The name of the Result Matrix. This can be a GPSS * \par * Matrix Entity up to 6 dimensions of any shape. Each * \par * Dimension represents either a factor in the experiment * \par * or replicate runs within a cell. * \par * 2. The dimension in the Matrix used for replicates. * \par * This argument can be 0, if no replicates are in the * \par * experimental design. * \par * 3. The Limit of Factor Interactions to be analyzed. * \par * This value can be 1, 2, or 3. It is normally used to * \par * preserve Degrees of Freedom for the estimate of the * \par * Standard Error instead of using them on an interaction * \par * presumed to be unimportant. * \par * * \par * Table of Residuals * \par * If you have previously defined a GPSS Table Entity of * \par * the name xxxx_Residuals, where xxxx is the name of the * \par * Result Matrix, the ANOVA Routine will automatically * \par * tabulate the residuals of the analysis. * \par * * \par * The Result Matrix * \par * Can be up to 6 dimensions. * \par * Each dimension used represents a factor in the experiment, * \par * with any number of treatment levels. Optionally, one of * \par * the dimensions (1-6) can be used for replicates within * \par * the experiment. This improves the estimate of the * \par * Standard Error. * \par * You can limit the level of factor interactions to be * \par * included in the analysis. Doing so provides more * \par * Degrees of Freedom to be used in the estimate of the * \par * Standard Error. * \par * * \par * Notes * \par * Use only CLEAR OFF during Experiments. CLEAR ON will give * \par * the value of 0.0 to all results. * \par * You should INITIALize the Result Matrix to UNSPECIFIED. * \par * Then you can perform ANOVA even when data are missing. * \par * You should define a Table Entity to receive the residuals * \par * of the ANOVA. * \par * To create an Experiment which can be saved and restarted, * \par * use Global Matrices or Savevalues to indicate which * \par * runs of the Experiment have been completed. Use these * \par * values as the starting point when the Experiment runs. * \par * Then you will be able to start an Experiment where at * \par * the end of the last completed run.\tab * \par * * \par *********************************************************************** \par \par * Define the Global Experimental Result Matrix * \par * Max of 3 Levels of Factor 1, 4 Levels of Factor 2, \par * Max of 3 Replicates per cell * \par MainResult MATRIX ,3,4,2 \par * Prepare the Result Matrix for ANOVA * \par * Initialize matrix elements to UNSPECIFIED instead of 0.0 \par INITIAL MainResult,UNSPECIFIED\tab \par * Prepare a Table for the residuals * \par MainResult_Residuals TABLE ,-5,1,10 \par \par \par PROCEDURE FillMatrix() BEGIN \par \par \tab MainResult[1,1,1] = 2; \par \tab MainResult[1,1,2] = 4; \par \tab MainResult[1,2,1] = 1; \par \tab MainResult[1,2,2] = 1; \par \tab MainResult[1,3,1] = 4; \par \tab MainResult[1,3,2] = 5; \par \tab MainResult[1,4,1] = 1; \par \tab MainResult[1,4,2] = 2; \par \tab MainResult[2,1,1] = 5; \par \tab MainResult[2,1,2] = 3; \par \tab MainResult[2,2,1] = 3; \par \tab MainResult[2,2,2] = 6; \par \tab MainResult[2,3,1] = 7; \par \tab MainResult[2,3,2] = 4; \par \tab MainResult[2,4,1] = 4; \par \tab MainResult[2,4,2] = 8; \par \tab MainResult[3,1,1] = 3; \par \tab MainResult[3,1,2] = 1; \par \tab MainResult[3,2,1] = 1; \par \tab MainResult[3,2,2] = 3; \par \tab MainResult[3,3,1] = 4; \par \tab MainResult[3,3,2] = 6; \par \tab MainResult[3,4,1] = 4; \par \tab MainResult[3,4,2] = 2; \par \par END; \par \par \par \par }