S-PLUS : Copyright (c) 1988, 1996 MathSoft, Inc. S : Copyright AT&T. Version 3.4 Release 1 for Sun SPARC, SunOS 5.3 : 1996 Working data will be in .Data y glyc.hemo log.duration age.diag dia.bp 1 3 13.7 2.332144 29.9 89 2 2 13.5 2.292535 18.6 90 3 3 13.8 2.747271 28.7 85 4 1 8.4 3.540959 20.1 99 5 3 12.8 3.131137 29.3 84 [1] "ncuts = 3 \n\n" id y.crm h.crm level Int1 Int2 Int3 glyc.hemo [1,] 1 0 1 1 1 0 0 13.7 [2,] 1 0 1 2 0 1 0 13.7 [3,] 1 1 1 3 0 0 1 13.7 [4,] 2 0 1 1 1 0 0 13.5 [5,] 2 1 1 2 0 1 0 13.5 [6,] 2 0 0 3 0 0 1 13.5 [7,] 3 0 1 1 1 0 0 13.8 [8,] 3 0 1 2 0 1 0 13.8 [9,] 3 1 1 3 0 0 1 13.8 [10,] 4 1 1 1 1 0 0 8.4 [11,] 4 0 0 2 0 1 0 8.4 [12,] 4 0 0 3 0 0 1 8.4 [13,] 5 0 1 1 1 0 0 12.8 [14,] 5 0 1 2 0 1 0 12.8 [15,] 5 1 1 3 0 0 1 12.8 id y level glyc.hemo log.duration age.diag dia.bp 1 1 0 : 1 13.7 2.332144 29.9 89 2 1 0 : 2 13.7 2.332144 29.9 89 3 1 1 : 3 13.7 2.332144 29.9 89 4 2 0 : 1 13.5 2.292535 18.6 90 5 2 1 : 2 13.5 2.292535 18.6 90 6 3 0 : 1 13.8 2.747271 28.7 85 7 3 0 : 2 13.8 2.747271 28.7 85 8 3 1 : 3 13.8 2.747271 28.7 85 9 4 1 : 1 8.4 3.540959 20.1 99 10 5 0 : 1 12.8 3.131137 29.3 84 11 5 0 : 2 12.8 3.131137 29.3 84 12 5 1 : 3 12.8 3.131137 29.3 84 13 6 0 : 1 13.0 3.258097 21.8 70 14 6 0 : 2 13.0 3.258097 21.8 70 15 6 1 : 3 13.0 3.258097 21.8 70 : 1 : 2 : 3 720 445 175 Call: glm(formula = y ~ glyc.hemo, family = binomial, data = wisc.crm.data, subset = ( as.integer(level) == 1)) Deviance Residuals: Min 1Q Median 3Q Max -1.130001 -0.9931404 -0.932452 1.35361 1.641739 Coefficients: Value Std. Error t value (Intercept) 0.22510208 0.38001813 0.5923456 glyc.hemo -0.05627671 0.02974703 -1.8918429 (Dispersion Parameter for Binomial family taken to be 1 ) Null Deviance: 957.6115 on 719 degrees of freedom Residual Deviance: 953.999 on 718 degrees of freedom Number of Fisher Scoring Iterations: 2 Call: glm(formula = y ~ glyc.hemo, family = binomial, data = wisc.crm.data, subset = ( as.integer(level) == 2)) Deviance Residuals: Min 1Q Median 3Q Max -1.477654 -1.347558 0.9541391 1.003686 1.161928 Coefficients: Value Std. Error t value (Intercept) 1.01503739 0.48744691 2.082355 glyc.hemo -0.04551101 0.03731596 -1.219613 (Dispersion Parameter for Binomial family taken to be 1 ) Null Deviance: 596.4632 on 444 degrees of freedom Residual Deviance: 594.9763 on 443 degrees of freedom Number of Fisher Scoring Iterations: 2 Call: glm(formula = y ~ glyc.hemo, family = binomial, data = wisc.crm.data, subset = ( as.integer(level) == 3)) Deviance Residuals: Min 1Q Median 3Q Max -1.885042 -1.530168 0.7670732 0.8080944 0.9018594 Coefficients: Value Std. Error t value (Intercept) 0.27807196 0.87839180 0.3165694 glyc.hemo 0.05636293 0.06755021 0.8343856 (Dispersion Parameter for Binomial family taken to be 1 ) Null Deviance: 203.6415 on 174 degrees of freedom Residual Deviance: 202.9303 on 173 degrees of freedom Number of Fisher Scoring Iterations: 3 Call: glm(formula = y ~ level + glyc.hemo, family = binomial, data = wisc.crm.data) Deviance Residuals: Min 1Q Median 3Q Max -1.709838 -1.004416 0.7307891 1.022713 1.568123 Coefficients: Value Std. Error t value (Intercept) 0.02245317 0.28434528 0.07896444 level: 2 0.92319834 0.12399549 7.44541891 level: 3 1.50031056 0.18751941 8.00082817 glyc.hemo -0.04008980 0.02184728 -1.83500205 (Dispersion Parameter for Binomial family taken to be 1 ) Null Deviance: 1857.608 on 1339 degrees of freedom Residual Deviance: 1754.334 on 1336 degrees of freedom Number of Fisher Scoring Iterations: 3 Analysis of Deviance Table Response: y Terms Resid. Df Resid. Dev Test Df Deviance 1 level + glyc.hemo 1336 1754.334 2 level * glyc.hemo 1334 1751.906 +level:glyc.hemo 2 2.428885