11 The SAS System 21:58 Tuesday, April 29, 2003 NOTE: Copyright (c) 1999-2001 by SAS Institute Inc., Cary, NC, USA. NOTE: SAS (r) Proprietary Software Release 8.2 (TS2M0) Licensed to GEORGE WASHINGTON UNIVERSITY, Site 0001180002. NOTE: This session is executing on the VM/ESA CMS Level 19 platform. NOTE: Running on IBM Model 7060 Serial Number 0106BA. 1 *** st210i sas; 2 CMS REL 492 (DET 2 ! ; 3 CMS CP LINK WORKVM 492 492 RR 3 ! ; 4 CMS ACC 492 f 4 ! ; 5 CMS FI NCGS DISK DUMMY NCGS f 5 ! ; 6 7 options formdlim='-'; 8 9 filename s_ilst 's_ilst pdf a'; 10 11 ods printer pdf file=s_ilst; NOTE: Writing ODS PRINTER output to DISK destination "S_ILST PDF A", printer "PDF". 12 13 DATA ALL; SET NCGS.ALLCHOL; 14 IF TRETGRP = 3 OR TRETGRP = 5; 15 y1=ch4; y2=ch6; y3=ch8; y4=ch9; 16 NOTE: There were 610 observations read from the data set NCGS.ALLCHOL. NOTE: The data set WORK.ALL has 610 observations and 43 variables. 17 proc sort; by ncgs; NOTE: There were 610 observations read from the data set WORK.ALL. NOTE: The data set WORK.ALL has 610 observations and 43 variables. 18 proc print; var ncgs pchol ch4 ch6 ch8 ch9; 19 NOTE: There were 610 observations read from the data set WORK.ALL. NOTE: The PROCEDURE PRINT printed pages 1-11. 20 PROC glm; CLASS tretgrp; 21 MODEL y1-y4 = tretgrp; 22 REPEATED y 4; 23 MANOVA H=tretgrp; 24 MEANS tretgrp; 25 LSMEANS tretgrp; 26 TITLE1 'REPEATED MEASURES ANALYSIS of cholesterol levels'; 27 title2 'Unbalanced ANOVA through GLM'; 28 29 RUN; NOTE: The PROCEDURE GLM printed pages 12-22. 12 The SAS System 21:58 Tuesday, April 29, 2003 30 DATA ONE; SET ALL; 31 KEEP NCGS TRETGRP X Y TIME XTIME; 32 X = PCHOL; 33 Y=CH4; TIME = 0; XTIME = X*TIME; OUTPUT; 34 Y=CH6; TIME = 6; XTIME = X*TIME; OUTPUT; 35 Y=CH8; TIME = 14; XTIME = X*TIME; OUTPUT; 36 Y=CH9; TIME = 18; XTIME = X*TIME; OUTPUT; 37 NOTE: There were 610 observations read from the data set WORK.ALL. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 38 data list; set; if _n_<50; NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.LIST has 49 observations and 6 variables. 39 proc print; 40 NOTE: There were 49 observations read from the data set WORK.LIST. NOTE: The PROCEDURE PRINT printed page 23. 41 data one; set one; NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 42 PROC REG; MODEL Y=X; 43 title2 'regression on baseline'; NOTE: 2440 observations read. NOTE: 483 observations have missing values. NOTE: 1957 observations used in computations. NOTE: The PROCEDURE REG printed page 24. 44 PROC REG; MODEL Y=X XTIME; 45 NOTE: 2440 observations read. NOTE: 483 observations have missing values. NOTE: 1957 observations used in computations. NOTE: The PROCEDURE REG printed page 25. 46 PROC SORT; BY TRETGRP; 47 NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 48 PROC MIXED METHOD=ML; 49 CLASS NCGS TRETGRP; 50 MODEL Y = TRETGRP X / S; 51 REPEATED / TYPE=UN SUBJECT=NCGS R; 52 lsmeans tretgrp / pdiff; 53 title2 'homogeneous mixed model analysis using maximum likelihood'; 54 title3 'unstructured covariance matrix'; 55 13 The SAS System 21:58 Tuesday, April 29, 2003 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 26-31. 56 PROC MIXED METHOD=ML; 57 CLASS NCGS TRETGRP; 58 MODEL Y = TRETGRP X / S; 59 REPEATED / TYPE=cs SUBJECT=NCGS R; 60 lsmeans tretgrp / pdiff; 61 title2 'homogeneous mixed model analysis using maximum likelihood'; 62 title3 'compound symmetry covariance matrix'; 63 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 32-37. 64 PROC MIXED METHOD=ML; 65 CLASS NCGS TRETGRP; 66 MODEL Y = TRETGRP TIME TIME*TRETGRP X X*TIME / S; 67 REPEATED / TYPE=UN SUBJECT=NCGS R; 68 lsmeans tretgrp / pdiff; 69 title2 'mixed model analysis with Quantitative time effects'; 70 title3 'with interactions'; 71 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 38-43. 72 PROC MIXED METHOD=ML; 73 CLASS NCGS TRETGRP; 74 MODEL Y = TRETGRP TIME X / S; 75 REPEATED / TYPE=UN SUBJECT=NCGS R; 76 lsmeans tretgrp / pdiff; 77 title3 'without interactions'; 78 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 44-49. 79 PROC MIXED METHOD=ML; 80 CLASS NCGS TRETGRP; 81 MODEL Y = TRETGRP TIME(TRETGRP) X / NOINT S; 82 REPEATED / TYPE=UN SUBJECT=NCGS R; 83 title3 'nested model'; 84 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 50-54. 85 PROC MIXED METHOD=ML; 86 CLASS NCGS TRETGRP time; 87 MODEL Y = TRETGRP TIME TIME*TRETGRP X X*TIME / S; 88 REPEATED / TYPE=UN SUBJECT=NCGS R; 14 The SAS System 21:58 Tuesday, April 29, 2003 89 lsmeans tretgrp time time*tretgrp/ pdiff; 90 title2 'mixed model analysis with Qualtitative time effects'; 91 title3 'with interactions'; 92 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 55-61. 93 proc sort; by ncgs; NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 94 proc reg outest=ncgsdata noprint; 95 model y = time; by ncgs; 96 title2 'naieve within subjects linear models'; 97 NOTE: 4 observations read. NOTE: 4 observations used in computations. NOTE: The above message was for the following by-group: NCGS=01002 NOTE: 4 observations read. NOTE: 4 observations used in computations. NOTE: The above message was for the following by-group: NCGS=01003 ********** :LINES DELETED NCGS=12077 NOTE: 4 observations read. NOTE: 1 observations have missing values. NOTE: 3 observations used in computations. NOTE: The above message was for the following by-group: NCGS=12078 NOTE: The data set WORK.NCGSDATA has 561 observations and 8 variables. 98 data test; merge all (keep = ncgs tretgrp) 99 ncgsdata (keep = ncgs intercept time); by ncgs; NOTE: There were 610 observations read from the data set WORK.ALL. NOTE: There were 561 observations read from the data set WORK.NCGSDATA. NOTE: The data set WORK.TEST has 610 observations and 4 variables. 100 proc print; NOTE: There were 610 observations read from the data set WORK.TEST. NOTE: The PROCEDURE PRINT printed pages 62-73. 101 proc univariate plot; var intercept time; 102 title2 'distribution of intercepts and slopes'; NOTE: The PROCEDURE UNIVARIATE printed pages 74-77. 151 The SAS System 21:58 Tuesday, April 29, 2003 103 proc ttest; class tretgrp; var intercept time; 104 title2 'naieve test of differences in intercepts and slopes'; 105 106 RUN; NOTE: There were 610 observations read from the data set WORK.TEST. NOTE: The PROCEDURE TTEST printed page 78. 107 108 data one; set one; NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 109 PROC MIXED METHOD=ML; 110 CLASS NCGS TRETGRP; 111 MODEL Y = TRETGRP TIME TIME*TRETGRP X / S; 112 random intercept time / type=un subject=ncgs; 113 lsmeans tretgrp / pdiff; 114 title2 'random effects slope model'; 115 title3 'ordinary maximum likelihood'; 116 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 79-84. 117 PROC MIXED METHOD=REML; 118 CLASS NCGS TRETGRP; 119 MODEL Y = TRETGRP TIME TIME*TRETGRP X / S; 120 random intercept time / type=un subject=ncgs; 121 title2 'random effects slope model'; 122 title3 'Restricted maximum likelihood'; 123 NOTE: 483 observations are not included because of missing values. NOTE: Convergence criteria met. NOTE: The PROCEDURE MIXED printed pages 85-89. 124 PROC SORT; BY TRETGRP time; 125 NOTE: There were 2440 observations read from the data set WORK.ONE. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 126 PROC genmod data=one; 127 CLASS ncgs TRETGRP; 128 MODEL Y = TRETGRP X time / dist=normal link=identity type3 waldci; 129 REPEATED SUBJECT=NCGS / sorted type=exch; 130 title2 'GEE marginal analysis'; 131 title3 'homogeneous marginal model'; 132 WARNING: Class levels for some variables were not printed due to excessive size. NOTE: Algorithm converged. NOTE: The scale parameter was estimated by maximum likelihood. NOTE: Algorithm converged. 152 The SAS System 21:58 Tuesday, April 29, 2003 NOTE: The PROCEDURE GENMOD printed pages 90-91. 133 PROC genmod data=one; 134 CLASS ncgs TRETGRP; 135 MODEL Y = TRETGRP X time tretgrp*x tretgrp*time x*time 136 / dist=normal link=identity type3 waldci; 137 REPEATED SUBJECT=NCGS / sorted type=exch; 138 title3 'test of interactions'; 139 WARNING: Class levels for some variables were not printed due to excessive size. NOTE: Algorithm converged. NOTE: The scale parameter was estimated by maximum likelihood. NOTE: Algorithm converged. NOTE: The PROCEDURE GENMOD printed pages 92-94. 140 DATA ONE; SET ALL; 141 KEEP NCGS TRETGRP X Y TIME XTIME; 142 X = PCHOL; 143 ih4=0; if (ch4>250) then ih4=1; 144 ih6=0; if (ch6>250) then ih6=1; 145 ih8=0; if (ch8>250) then ih8=1; 146 ih9=0; if (ch9>250) then ih9=1; 147 Y=iH4; TIME = 0; XTIME = X*TIME; OUTPUT; 148 Y=iH6; TIME = 6; XTIME = X*TIME; OUTPUT; 149 Y=iH8; TIME = 14; XTIME = X*TIME; OUTPUT; 150 Y=iH9; TIME = 18; XTIME = X*TIME; OUTPUT; 151 NOTE: There were 610 observations read from the data set WORK.ALL. NOTE: The data set WORK.ONE has 2440 observations and 6 variables. 152 PROC genmod data=one descending; 153 CLASS ncgs TRETGRP; 154 MODEL Y = TRETGRP X / dist=bin link=logit type3 waldci; 155 REPEATED SUBJECT=NCGS / sorted type=exch; 156 title1 'GEE marginal logistic regression analysis'; 157 title2 'odds of elevated cholesterol > 250 mg/DL'; 158 WARNING: Class levels for some variables were not printed due to excessive size. NOTE: PROC GENMOD is modeling the probability that Y='1'. NOTE: Algorithm converged. NOTE: The scale parameter was held fixed. NOTE: Algorithm converged. NOTE: The PROCEDURE GENMOD printed pages 95-96. 159 PROC genmod data=one descending; 160 CLASS ncgs TRETGRP; 161 MODEL Y = TRETGRP X time tretgrp*x tretgrp*time x*time 162 / dist=bin link=logit type3 waldci; 163 REPEATED SUBJECT=NCGS / sorted type=exch; 164 title3 'interaction model'; 165 166 RUN; WARNING: Class levels for some variables were not printed due to excessive size. NOTE: PROC GENMOD is modeling the probability that Y='1'. 153 The SAS System 21:58 Tuesday, April 29, 2003 NOTE: Algorithm converged. NOTE: The scale parameter was held fixed. NOTE: Algorithm converged. NOTE: The PROCEDURE GENMOD printed pages 97-99. 167 168 ods printer close; NOTE: ODS PRINTER printed 104 pages to S_ILST PDF A. 169 170 CMS REL f (DET 170 ! ; ERROR: Errors printed on pages 5,6,8,10,11,12,14,15,16,19,20,21,22,24,26,27,30,31,32,33,34,35,36,37,40,41,44,45,46,47,48,50. +ERROR: Errors printed on pages 5,6,8,10,11,12,14,15,16,19,20,21,22,24,26,27,30,31,32,33,34,35,36,37,40,41,44,45,46,47,48,50. +ERROR: Errors printed on pages 5,6,8,10,11,12,14,15,16,19,20,21,22,24,26,27,30,31,32,33,34,35,36,37,40,41,44,45,46,47,48,50. NOTE: SAS Institute Inc., SAS Campus Drive, Cary, NC USA 27513-2414