The RELIABILITY Procedure

Comparison of Two Samples of Repair Data

Nelson (2002) and Doganaksoy and Nelson (1998) show how the difference of MCFs from two samples can be used to compare the populations from which they are drawn. The RELIABILITY procedure provides Doganaksoy and Nelson’s confidence intervals for the pointwise difference of the two MCFs, which can be used to assess whether the difference is statistically significant.

Doganaksoy and Nelson (1998) give an example of two samples of locomotives with braking grids from two different production batches. Figure 16.35 contains a listing of the data. The variable ID is a unique identifier for individual locomotives. The variable Days provides the locomotive age in days. The variable Value is 1 if the age corresponds to a valve seat replacement or -1 if the age corresponds to the locomotive’s latest age (the current end of its history). The variable Sample is a group variable that identifies the grid production batch.

data Grids;                                                    
   if _N_ < 40 then Sample = 'Sample1';                       
   else Sample = 'Sample2';                                   
   input ID$ Days Value @@;                                   
   datalines;                                                     
S1-01 462  1    S1-01 730 -1    S1-02 364  1    S1-02 391  1  
S1-02 548  1    S1-02 724 -1    S1-03 302  1    S1-03 444  1  
S1-03 500  1    S1-03 730 -1    S1-04 250  1    S1-04 730 -1  
S1-05 500  1    S1-05 724 -1    S1-06  88  1    S1-06 724 -1  
S1-07 272  1    S1-07 421  1    S1-07 552  1    S1-07 625  1  
S1-07 719 -1    S1-08 481  1    S1-08 710 -1    S1-09 431  1  
S1-09 710 -1    S1-10 367  1    S1-10 710 -1    S1-11 635  1  
S1-11 650  1    S1-11 708 -1    S1-12 402  1    S1-12 700 -1  
S1-13  33  1    S1-13 687 -1    S1-14 287  1    S1-14 687 -1  
S1-15 317  1    S1-15 498  1    S1-15 657 -1    S2-01 203  1  
S2-01 211  1    S2-01 277  1    S2-01 373  1    S2-01 511 -1  
S2-02 293  1    S2-02 503 -1    S2-03 173  1    S2-03 470 -1  
S2-04 242  1    S2-04 464 -1    S2-05  39  1    S2-05 464 -1  
S2-06  91  1    S2-06 462 -1    S2-07 119  1    S2-07 148  1  
S2-07 306  1    S2-07 461 -1    S2-08 382  1    S2-08 460 -1  
S2-09 250  1    S2-09 434 -1    S2-10 192  1    S2-10 448 -1  
S2-11 369  1    S2-11 448 -1    S2-12  22  1    S2-12 447 -1  
S2-13  54  1    S2-13 441 -1    S2-14 194  1    S2-14 432 -1  
S2-15  61  1    S2-15 419 -1    S2-16  19  1    S2-16 185  1  
S2-16 419 -1    S2-17 187  1    S2-17 416 -1    S2-18  93  1  
S2-18 205  1    S2-18 264  1    S2-18 415 -1                  
;                                                             

Figure 16.35: Partial Listing of the Braking Grids Data

Obs Sample ID Days Value
1 Sample1 S1-01 462 1
2 Sample1 S1-01 730 -1
3 Sample1 S1-02 364 1
4 Sample1 S1-02 391 1
5 Sample1 S1-02 548 1
6 Sample1 S1-02 724 -1
7 Sample1 S1-03 302 1
8 Sample1 S1-03 444 1
9 Sample1 S1-03 500 1
10 Sample1 S1-03 730 -1
11 Sample1 S1-04 250 1
12 Sample1 S1-04 730 -1
13 Sample1 S1-05 500 1
14 Sample1 S1-05 724 -1
15 Sample1 S1-06 88 1
16 Sample1 S1-06 724 -1
17 Sample1 S1-07 272 1
18 Sample1 S1-07 421 1
19 Sample1 S1-07 552 1
20 Sample1 S1-07 625 1


The following statements request the Nelson (1995) nonparametric estimate and confidence limits for the difference of the MCF functions shown in Figure 16.36 for the braking grids:

proc reliability data=Grids;
   unitid ID;
   mcfplot Days*Value(-1) = Sample / mcfdiff;
run;

The MCFPLOT statement requests a plot of each MCF estimate as a function of age (provided by Days), and it specifies that the end of history for each system is identified by Value equal to -1. The variable Sample identifies the two samples of braking grids. The option MCFDIFF requests that the difference between the MCFs of the two groups given in the variable Sample be computed and plotted. Confidence limits for the MCF difference are also computed and plotted. The UNITID statement specifies that the variable Id uniquely identify each system.

Figure 16.36 shows the plot of the MCF difference function and pointwise 95% confidence intervals. Since the pointwise confidence limits do not include zero for some system ages, the difference between the two populations is statistically significant. A listing of the tabular output is shown in Figure 16.37. It contains a summary of the repair data for the two samples, estimates, standard errors, and confidence intervals for the MCF difference. A statistical test for different MCFs is also computed and is displayed in the table Tests for Equality of Mean Functions. The tests also indicate a significant difference between the two samples.

Figure 16.36: Mean Cumulative Function Difference

Mean Cumulative Function Difference


Figure 16.37: Listing of the Output for the Braking Grids Data

MCF Difference Data Summary
Input Data Set WORK.GRIDS
Group 1 Sample1
Observations Used 39
Number of Units 15
Number of Events 24
Group 2 Sample2
Observations Used 44
Number of Units 18
Number of Events 26

Sample MCF Differences
Age MCF Difference Standard Error 95% Confidence Limits Unit ID
Lower Upper
19.00 -0.056 0.054 -0.161 0.050 S2-16
22.00 -0.111 0.074 -0.256 0.034 S2-12
33.00 -0.044 0.098 -0.237 0.148 S1-13
39.00 -0.100 0.109 -0.313 0.113 S2-05
54.00 -0.156 0.117 -0.385 0.074 S2-13
61.00 -0.211 0.124 -0.453 0.031 S2-15
88.00 -0.144 0.137 -0.414 0.125 S1-06
91.00 -0.200 0.142 -0.478 0.078 S2-06
93.00 -0.256 0.145 -0.539 0.028 S2-18
119.00 -0.311 0.146 -0.598 -0.024 S2-07
148.00 -0.367 0.167 -0.693 -0.040 S2-07
173.00 -0.422 0.166 -0.748 -0.097 S2-03
185.00 -0.478 0.182 -0.835 -0.120 S2-16
187.00 -0.533 0.180 -0.886 -0.181 S2-17
192.00 -0.589 0.177 -0.935 -0.243 S2-10
194.00 -0.644 0.172 -0.982 -0.307 S2-14
203.00 -0.700 0.167 -1.027 -0.373 S2-01
205.00 -0.756 0.178 -1.105 -0.407 S2-18
211.00 -0.811 0.188 -1.179 -0.443 S2-01
242.00 -0.867 0.180 -1.219 -0.514 S2-04
250.00 -0.856 0.179 -1.207 -0.504 S1-04,S2-09
264.00 -0.911 0.202 -1.307 -0.515 S2-18
272.00 -0.844 0.208 -1.252 -0.437 S1-07
277.00 -0.900 0.227 -1.345 -0.455 S2-01
287.00 -0.833 0.231 -1.286 -0.380 S1-14
293.00 -0.889 0.222 -1.323 -0.455 S2-02
302.00 -0.822 0.224 -1.262 -0.383 S1-03
306.00 -0.878 0.241 -1.350 -0.406 S2-07
317.00 -0.811 0.242 -1.286 -0.337 S1-15
364.00 -0.744 0.242 -1.219 -0.270 S1-02
367.00 -0.678 0.241 -1.150 -0.206 S1-10
369.00 -0.733 0.230 -1.185 -0.282 S2-11
373.00 -0.789 0.257 -1.293 -0.284 S2-01
382.00 -0.844 0.246 -1.327 -0.362 S2-08
391.00 -0.778 0.261 -1.290 -0.266 S1-02
402.00 -0.711 0.258 -1.217 -0.206 S1-12
421.00 -0.644 0.270 -1.174 -0.115 S1-07
431.00 -0.578 0.265 -1.097 -0.059 S1-09
444.00 -0.511 0.275 -1.049 0.027 S1-03
462.00 -0.444 0.267 -0.968 0.079 S1-01
481.00 -0.378 0.258 -0.883 0.128 S1-08
498.00 -0.311 0.265 -0.830 0.208 S1-15
500.00 -0.244 0.253 -0.741 0.252 S1-05
500.00 -0.178 0.275 -0.716 0.360 S1-03

Tests for Equality of Mean Functions
Weight Function Statistic Variance Chi Square DF Pr > Chi Square
Constant -3.673285 4.556053 2.961560 1 0.0853
Linear -4.435032 1.424770 13.805393 1 0.0002


You can fit a parametric model that uses Sample as a classification variable. This results in a model with a common shape parameter for the two groups but with different scale parameters. Suppose you want estimates of the parametric mean and intensity functions at values of the time variable 500, 600, 700, 800, 900, and 1,000 days for each of the two groups. The following statements create a new input data set that has observations for the desired prediction times appended to it. The additional observations are not used in the analysis, because the censoring variable Value is set to missing for those observations. Values of the mean and intensity function are computed, however, in the table that is produced by specifying the OBSTATS option in the MODEL statement.

The following statements create the new data set by appending observations to the original Grids data set:

data Predict;
   Control=1;
   if _N_ < 7 then Sample = 'Sample1';
   else Sample = 'Sample2';

   input ID$ Days Value;
   cards;
9999 500  .
9999 600  .   
9999 700  .
9999 800  .
9999 900  .
9999 1000 .
9999 500  .
9999 600  .   
9999 700  .
9999 800  .
9999 900  .
9999 1000 .
;

data Grids;
   set Predict Grids;
run;

The following statements fit a nonhomogeneous Poisson process with a power law mean function that uses Sample as a two-level covariate. The OBSTATS option requests that predicted values be computed for values of the variable Control equal to 1. The MCFPLOT statement plots the fitted model as well as the nonparametric estimates of the MCF. Parametric confidence limits are displayed by default.

proc reliability data=Grids;
   unitid ID;
   distribution nhpp(pow);
   class Sample;
   model   Days*Value(-1) = Sample /obstats(control=Control);
   mcfplot Days*Value(-1) = Sample /fit=model overlay;
run;

Figure 16.38: Predicted Mean and Intensity Function for the Braking Grids Data

The RELIABILITY Procedure

Observation Statistics
Days Value Sample ID Xbeta Shape MCF MCF_Lower MCF_Upper MCF_StdErr Intensity Int_Lower Int_Upper Int_StdErr
500 . Sample1 9999 464.04647 1.1050556 1.0859585 0.7176451 1.6432995 0.2295199 0.0024001 0.0015543 0.0037061 0.000532
600 . Sample1 9999 464.04647 1.1050556 1.3283512 0.8874029 1.9884054 0.2733977 0.0024465 0.0015458 0.0038719 0.0005731
700 . Sample1 9999 464.04647 1.1050556 1.5750445 1.0556772 2.3499278 0.3215248 0.0024864 0.0015325 0.0040343 0.000614
800 . Sample1 9999 464.04647 1.1050556 1.8254803 1.2215469 2.7279987 0.3741606 0.0025216 0.0015171 0.0041911 0.0006537
900 . Sample1 9999 464.04647 1.1050556 2.0792348 1.3846219 3.1223089 0.4313142 0.002553 0.0015011 0.0043418 0.0006917
1000 . Sample1 9999 464.04647 1.1050556 2.3359745 1.5448083 3.5323327 0.4928631 0.0025814 0.0014852 0.0044866 0.000728
500 . Sample2 9999 323.23791 1.1050556 1.6193855 1.1012136 2.3813813 0.3186232 0.003579 0.0021872 0.0058566 0.0008993
600 . Sample2 9999 323.23791 1.1050556 1.9808425 1.335699 2.9375905 0.3982653 0.0036482 0.0021492 0.0061928 0.0009849
700 . Sample2 9999 323.23791 1.1050556 2.3487125 1.5633048 3.5287107 0.4878045 0.0037078 0.0021124 0.006508 0.0010643
800 . Sample2 9999 323.23791 1.1050556 2.7221634 1.7845232 4.1524669 0.5864921 0.0037602 0.002078 0.0068042 0.0011378
900 . Sample2 9999 323.23791 1.1050556 3.100563 2.0000301 4.8066733 0.6935604 0.003807 0.002046 0.0070837 0.0012061
1000 . Sample2 9999 323.23791 1.1050556 3.4834144 2.2104841 5.4893747 0.8083117 0.0038494 0.0020164 0.0073484 0.0012699


Figure 16.39: Fitted Model

Fitted Model


The predicted values of the mean and intensity functions at the desired values of Days, with standard errors and confidence limits, are shown in Figure 16.38.

A plot of the fitted mean function, along with nonparametric estimates for the two samples, is shown in Figure 16.39.