The MVPMONITOR Procedure

Example 13.4 Creating a Classical $T^2$ Chart

The following statements use PROC MVPMODEL to create a model from which classical $T^2$ charts can be produced:

proc mvpmodel data=flightDelays ncomp=all noprint out=mvpout;
   var AA CO DL F9 FL NW UA US WN;
run;

The NCOMP=ALL option specifies that the number of principal components equal the number of process variables, so the mvpout data set contains the classical $T^2$ statistic for each observation. The mvpout data set contains six observations per time point—one for each region.

The following statements create the classical $T^2$ chart:

proc mvpmonitor history=mvpout;
   time flightDate;
   series region;
   tsquarechart / seriesvalue='NE';
run;

The SERIES statement specifies region as the variable that identifies sequences of $T^2$ statistics, and the SERIESVALUE= option selects the region to be plotted, the Northeast. The classical $T^2$ chart is shown in Output 13.4.1.

Output 13.4.1: Classical $T^2$ Chart

Classical T2 Chart


In this case, the classical $T^2$ chart finds out-of-control observations above the upper control limit during February 14–16 and below the lower control limit on February 1, 10, and 12.

Output 13.4.2 shows a partial listing of the mvpout data set. It contains $T^2$ statistics based on the model that has nine principal components, in addition to the original variables and other observationwise statistics.

Output 13.4.2: Partial Listing of Output Data Set mvpout

flightDate region AA CO DL F9 FL NW UA US WN Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Prin9 _NOBS_ _TSQUARE_ R_AA R_CO R_DL R_F9 R_FL R_NW R_UA R_US R_WN _SPE_
02/01/07 MW 14.9 7.1 7.9 8.5 14.8 4.5 5.1 13.4 5.1 -1.16440 0.89425 0.00118 -0.84496 0.07443 -0.05231 -0.27403 0.31880 -0.06726 96 3.2122 0 0 0 0 0 0 0 0 0 .
02/01/07 NE 15.7 7.1 8.6 6.3 14.6 6.2 7.0 11.0 6.4 -1.08197 0.55936 0.05859 -0.84487 -0.02092 -0.16851 -0.30596 0.25090 0.12804 96 2.7349 0 0 0 0 0 0 0 0 0 .
02/01/07 NW 17.8 2.6 6.1 28.8 11.6 6.1 11.6 27.3 3.7 -0.31223 2.23593 -0.03824 0.58893 0.21628 0.67149 0.85708 1.07348 -0.49530 96 18.5555 0 0 0 0 0 0 0 0 0 .
02/01/07 SC 19.9 8.3 13.9 4.9 25.8 15.3 9.0 15.1 12.8 0.22974 -0.16976 0.56118 -1.30139 0.22990 -0.34406 0.11471 0.30201 0.46481 96 6.3335 0 0 0 0 0 0 0 0 0 .
02/01/07 SE 16.1 1.9 8.7 8.7 15.1 18.3 4.0 10.4 6.5 -0.87056 0.01044 1.06975 -0.50179 0.38041 -0.33789 0.63114 0.45028 0.21437 96 6.1537 0 0 0 0 0 0 0 0 0 .


Notice that no SPE statistics are produced when the number of principal components equals the number of process variables.