Description

Replace=FULL Radius=0 Maxclusters=2 Maxiter=20 Converge=0.02
Initial Seeds
Cluster x1 x2 x3 x4 x5 x6 x7 x8 x9 x10
1 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000 1.000000000
2 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000 5.000000000
Minimum Distance Between Initial Seeds = 13.85641
Iteration History
Iteration Criterion Relative Change in Cluster Seeds
1 2
1 1.4646 0.2369 0.2997
2 0.9506 0.0354 0.0186
3 0.9422 0.0175 0.00933
Cluster Means
Cluster x1 x2 x3 x4 x5 x6 x7 x8 x9 x10
1 1.924528302 1.819047619 1.730769231 1.688679245 1.634615385 1.613207547 1.641509434 1.692307692 1.716981132 1.886792453

Convergence criterion is satisfied.
Criterion Based on Final Seeds = 0.9407
Cluster Summary
Cluster Frequency RMS Std Deviation Maximum Distance from Seed to Observation Radius Exceeded Nearest Cluster Distance Between Cluster Centroids
1 107 1.0120 7.3426 2 8.1204
2 186 0.9024 6.5845 1 8.1204
Statistics for Variables
Variable Total STD Within STD R-Square RSQ/(1-RSQ)
x1 1.38836 1.36137 0.041821 0.043646
x2 1.34963 1.28146 0.101601 0.113092
x3 1.31256 1.13478 0.255149 0.342550
x4 1.36519 1.04190 0.419559 0.722828
x5 1.38784 0.91039 0.571195 1.332061
x6 1.43532 0.82200 0.673155 2.059552
x7 1.49185 0.74642 0.750532 3.008525
x8 1.53485 0.70022 0.792596 3.821498
x9 1.57357 0.65412 0.827799 4.807177
x10 1.57250 0.68648 0.810078 4.265312
x11 1.59060 0.77151 0.765549 3.265277
x12 1.60994 0.88004 0.702235 2.358356
OVER-ALL 1.47117 0.94401 0.589688 1.437172
Pseudo F Statistic = 418.22
Approximate Expected Over-All R-Squared = 0.08070
Cubic Clustering Criterion = 197.210

2 2.513513514 2.711956522 3.108695652 3.521739130 3.816216216 4.054347826 4.320652174 4.530054645 4.684782609 4.820652174
Cluster Standard Deviations
Cluster x1 x2 x3 x4 x5 x6 x7 x8 x9 x10
1 1.357154679 1.158421940 1.026169891 0.919246146 0.738112159 0.724677330 0.770759418 0.813441963 0.870191188 1.017192726
2 1.363770780 1.346384951 1.191559769 1.106150326 0.993870758 0.872959046 0.732096032 0.627154307 0.488766661 0.384690080
Distance Between Cluster Centroids
Nearest Cluster 1 2
1 . 8.120413511
2 8.120413511 .

The LOGISTIC Procedure
Model Information
Data Set WORK.TMP
Response Variable CLUSTER Cluster
Number of Response Levels 2
Model binary logit
Optimization Technique Fisher’s scoring
Number of Observations Read 243
Number of Observations Used 236
Response Profile
Ordered Value CLUSTER Total Frequency
1 1 84
2 2 152
Probability modeled is CLUSTER=1.
Model Convergence Status
Convergence criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Criterion Intercept Only Intercept and Covariates
AIC 309.292 322.554
SC 312.756 374.512
-2 Log L 307.292 292.554
Testing Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood Ratio 14.7373 14 0.3963
Score 14.7006 14 0.3989
Wald 13.2554 14 0.5065
Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Standard Error Wald Chi-Square Pr > ChiSq
Intercept 1 -1.0579 1.5415 0.4710 0.4925
dist 1 -0.0988 0.1030 0.9211 0.3372
Age 1 0.0984 0.1543 0.4067 0.5237
Gender 1 -0.0256 0.3530 0.0053 0.9422
Married 1 -0.1772 0.4044 0.1920 0.6612
License 1 0.5203 0.9234 0.3175 0.5731
Adults 1 -0.3329 0.2522 1.7424 0.1868
Children 1 0.0246 0.1757 0.0195 0.8888
Cars 1 0.4054 0.2196 3.4069 0.0649
Education 1 -0.0881 0.1061 0.6905 0.4060
SpouseEd 1 -0.1225 0.1213 1.0208 0.3123
Years 1 -0.0330 0.1212 0.0740 0.7855
Workers 1 -0.0913 0.2300 0.1574 0.6915
Income 1 -0.0247 0.0930 0.0705 0.7907

Ethnic 1 1.0567 0.5011 4.4462 0.0350
Odds Ratio Estimates
Effect Point Estimate 95% Wald Confidence Limits
dist 0.906 0.740 1.108
Age 1.103 0.815 1.493
Gender 0.975 0.488 1.947
Married 0.838 0.379 1.850
License 1.682 0.275 10.278
Adults 0.717 0.437 1.175
Children 1.025 0.726 1.446
Cars 1.500 0.975 2.307
Education 0.916 0.744 1.127
SpouseEd 0.885 0.698 1.122
Years 0.968 0.763 1.227
Workers 0.913 0.582 1.433
Income 0.976 0.813 1.171
Ethnic 2.877 1.077 7.682
Association of Predicted Probabilities and Observed Responses
Percent Concordant 64.4 Somers’ D 0.293
Percent Discordant 35.1 Gamma 0.295
Percent Tied 0.5 Tau-a 0.135
Pairs 12768 c 0.647

Actual vs. Predicted on hold-out sample using LR
The FREQ Procedure

Frequency
Percent
Row Pct
Col Pct

Table of actual by into_l
actual into_l
1 2 Total
1 4
7.02
17.39
57.14 19
33.33
82.61
38.00 23
40.35
2 3
5.26
8.82
42.86 31
54.39
91.18
62.00 34
59.65
Total 7
12.28 50
87.72 57
100.00

Total Sample Size 236 DF Total 235
Variables 14 DF Within Classes 234
Classes 2 DF Between Classes 1
Number of Observations Read 243
Number of Observations Used 236
Class Level Information
CLUSTER Variable
Name Frequency Weight Proportion Prior Probability
1 _1 84 84.0000 0.355932 0.500000
2 _2 152 152.0000 0.644068 0.500000
Pooled Covariance Matrix Information
Covariance Matrix Rank Natural Log of the
Determinant of the Covariance Matrix
14 -8.67171

Generalized Squared Distance to
CLUSTER
From CLUSTER 1 2
1 0 0.28732
2 0.28732 0

Canonical Discriminant Analysis
t
Canonical Correlation Adjusted
Canonical
Correlation Approximate
Standard Error Squared
Canonical Correlation Eigenvalues of Inv(E)*H = CanRsq/(1-CanRsq) Likelihood Ratio Test of H0: The canonical correlations in
Approximat
F Valu
Eigenvalue Difference Proportion Cumulative
1 0.249581 0.146158 0.061169 0.062291 0.0664 1.0000 1.0000 0.93770928 1.0
e e 5
Note: The F statistic is exact.
Canonical Discriminant Analysis
Total Canonical Structure
Variable Label Can1
dist dist 0.222276
Age Age -0.162976
Gender Gender -0.017887
Married Married 0.125933
License License -0.106278
Adults Adults 0.089158
Children Children -0.105804
Cars Cars -0.265394
Education Education 0.431749
SpouseEd SpouseEd 0.473609
Years Years 0.043631
Workers Workers 0.069582
Income Income 0.198403
Ethnic Ethnic -0.590094
Between Canonical Structure
Variable Label Can1
dist dist 1.000000
Age Age -1.000000
Gender Gender -1.000000
Married Married 1.000000
License License -1.000000
Adults Adults 1.000000
Children Children -1.000000
Cars Cars -1.000000
Education Education 1.000000
SpouseEd SpouseEd 1.000000
Years Years 1.000000
Workers Workers 1.000000
Income Income 1.000000
Ethnic Ethnic -1.000000
Pooled Within Canonical Structure
Variable Label Can1
dist dist 0.215573
Age Age -0.157949
Gender Gender -0.017321
Married Married 0.122008
License License -0.102950
Adults Adults 0.086358
Children Children -0.102491
Cars Cars -0.257561
Education Education 0.420534
SpouseEd SpouseEd 0.461859
Years Years 0.042253
Workers Workers 0.067390
Income Income 0.192360
Ethnic Ethnic -0.577719

Page 11 of 1
Canonical Discriminant Analysis
Total-Sample Standardized Canonical
Coefficients
Variable Label Can1
dist dist 0.2639731240
Age Age -.2123378243
Gender Gender 0.0317802985
Married Married 0.1450579939
License License -.1683997920
Adults Adults 0.5295221319
Children Children -.0531030230
Cars Cars -.7539062621
Education Education 0.2707008680
SpouseEd SpouseEd 0.3049188801
Years Years 0.0806570008
Workers Workers 0.1245973956
Income Income 0.0972594872
Ethnic Ethnic -.6031578697
Pooled Within-Class Standardized Canonical
Coefficients
Variable Label Can1
dist dist 0.2641291903
Age Age -.2126149478
Gender Gender 0.0318478154
Married Married 0.1452957966
License License -.1686998598
Adults Adults 0.5305209883
Children Children -.0531978127
Cars Cars -.7538562703
Education Education 0.2696991083
SpouseEd SpouseEd 0.3034274798
Years Years 0.0808243684
Workers Workers 0.1248445152
Income Income 0.0973475170
Ethnic Ethnic -.5978540616
Raw Canonical Coefficients
Variable Label Can1
dist dist 0.181670146
Age Age -0.187772908
Gender Gender 0.067417409
Married Married 0.322770738
License License -0.990523360
Adults Adults 0.637298225
Children Children -0.059038883
Cars Cars -0.758467223
Education Education 0.170465203
SpouseEd SpouseEd 0.234075906
Years Years 0.059623606
Workers Workers 0.163004461
Income Income 0.046805254

Ethnic Ethnic -2.051929931

Class Means on Canonical Variables
CLUSTER Can1
1 -.3452324822
2 0.1907863718

Linear Discriminant Function for CLUSTER
Variable Label 1 2
Constant -60.94135 -60.42816
dist dist 1.19085 1.28823
Age Age 4.25007 4.14942
Gender Gender 4.54253 4.57866
Married Married 10.55040 10.72341
License License 32.93791 32.40697
Adults Adults -0.13306 0.20854
Children Children 2.43716 2.40551
Cars Cars 3.45684 3.05029
Education Education 2.07917 2.17054
SpouseEd SpouseEd 0.79410 0.91957
Years Years 0.94753 0.97949
Workers Workers 0.73014 0.81751
Income Income 1.22045 1.24554
Ethnic Ethnic 12.61094 11.51107

Classification Summary for Calibration Data: WORK.TMP Resubstitution Summary using Linear Discriminant Function
Number of Observations and Percent Classified into CLUSTER
From CLUSTER 1 2 Total
. 3
42.86 4
57.14 7
100.00
1 44
52.38 40
47.62 84
100.00
2 47
30.92 105
69.08 152
100.00
Total 94
38.68 149
61.32 243
100.00
Priors 0.5 0.5
Error Count Estimates for CLUSTER
1 2 Total
Rate 0.4762 0.3092 0.3927
Priors 0.5000 0.5000

Classification Summary for Test Data: WORK.TMP_HOLD Classification Summary using Linear Discriminant Function
Observation Profile for Test Data
Number of Observations Read 57
Number of Observations Used 57
Number of Observations and Percent Classified into CLUSTER
1 2 Total
Total 22
38.60 35
61.40 57
100.00
Priors 0.5 0.5

Actual vs. Predicted on hold-out sample using DA
The FREQ Procedure

Frequency
Percent
Row Pct
Col Pct

Table of actual by _INTO_
actual _INTO_(Cluster)
1 2 Total
1 11
19.30
47.83
50.00 12
21.05
52.17
34.29 23
40.35
2 11
19.30
32.35
50.00 23
40.35
67.65
65.71 34
59.65
Total 22
38.60 35
61.40 57
100.00

Prediction of DA vs. Prediction of LR on hold-out sample
The FREQ Procedure

Frequency
Percent
Row Pct
Col Pct

Table of _INTO_ by into_l
_INTO_(Cluster) into_l
1 2 Total
1 7
12.28
31.82 100.00 15
26.32
68.18
30.00 22
38.60
2 0
0.00
0.00
0.00 35
61.40
100.00
70.00 35
61.40
Total 7
12.28 50
87.72 57
100.00