Hi.
Consider a 2-year trial to compare 17 treatments. The treatments were observed in randomized complete block design with 3 block was used. Both year and the block are random effects. Test in each year did in one different location
I use this analysis method:
title;
footnote;
*** Factorial ANOVA ***;
data=work.test2;
input year location treat rep Cycloxidim1 Arrat Cycloxidim2 OHC NLATOT WLATOT BY SY;
cards;
.
.
.
;
proc glm data=Work.Test2;
class year location treat rep;
model NLATOT WLATOT BY SY = year rep treat year*rep treat*year treat*rep /tolerance solution;
random year rep treat year*rep treat*year treat*rep / test;
contrast 'new herbicide vs Ordinal Herbicide' treat 0 -5 2 2 2 2 2 2 2 2 2 2 0 -5 -5 -5 0;
contrast 'ordinal herbicide vs control' treat -4 1 0 0 0 0 0 0 0 0 0 0 1 1 1 0;
contrast 'new herbicide vs control' treat -11 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0;
contrast 'new herbicide vs weeding' treat 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 -11;
contrast 'ordinal herbicide vs weeding' treat 0 1 0 0 0 0 0 0 0 0 0 0 1 1 1 -4;
contrast 'new herbicide high dosage vs new herbicide low dosage' treat 0 0 1 1 -1 -1 1 1 -1 -1 0 0 0 0 0 0 0;
contrast 'new herbicide one treat vs new herbicide two treat' treat 0 0 1 1 1 1 -1 -1 -1 -1 0 0 0 0 0 0 0;
contrast 'arrat low dosage vs arrat high dosage' treat 0 0 1 1 0 0 0 0 0 0 -1 -1 0 0 0 0 0;
estimate 'year' year -1 1;
estimate 'block' Rep -1 1;
estimate 'treatment' Treat -1 1;
estimate 'year*rep ' year*rep 1 -1 -1 1 / DIVISOR=2;
estimate 'year*treat ' year*treat 1 -1 -1 1 / DIVISOR=2;
estimate 'Rep*treat' Rep*treat 1 -1 -1 1 / DIVISOR=2;
means year rep treat year*rep treat*year treat*rep / hovtest=bartlett;
means year rep treat year*rep treat*year treat*rep / DUNCAN;
output out=diag P=pred R=resid;
lsmeans year rep treat year*rep treat*year treat*rep;
test h=year rep e= treat*year;
run;
PROC UNIVARIATE DATA=diag NORMAL ;
VAR resid;
run;
PROC RSREG DATA=Work.Test2 PLOTS=(DIAGNOSTICS RIDGE);
MODEL BY SY = year treat rep / LACKFIT VIF TOL XPX COVB;
RIDGE MAX MIN;
run;
PROC REG DATA=Work.Test2 outest=est1 outsscp=sscp1 rsquare;
Title'PP Plot'
MODEL BY SY = NLATOT WLATOT cycloxidim1/ noprint;
Plot npp. *r. / annotate=test2 nostat cframe=ligr
modellab="'Best'Two-Parameter Model:";
run;
Title'QQ Plot'
Plot r. *nqq. / noline mse cframe=ligr
modellab="'Best'Two-Parameter Model:";
run;
PROC Corr DATA=Work.Test2;
Var BY SY NLATOT WLATOT cycloxidim1;
run;
and I see these problem in log window:
NOTE: Starting the Analyst Application.
.
.
.
NOTE: TYPE I EMS not available without the E1 option.
NOTE: year*rep is not estimable.
NOTE: year*treat is not estimable.
NOTE: Rep*treat is not estimable.
NOTE: Means from the MEANS statement are not adjusted for other terms in the model. For adjusted means, use the LSMEANS statement.
WARNING: Homogeneity of variance testing and Welch's ANOVA are only available for unweighted one-way models.
NOTE: The data set WORK.DIAG has 102 observations and 127 variables.
NOTE: PROCEDURE GLM used (Total process time):
real time 4.23 seconds
cpu time 1.12 seconds
.
.
35 PROC RSREG DATA=Work.Test2 PLOTS=(DIAGNOSTICS RIDGE);
-----
22
76
ERROR 22-322: Syntax error, expecting one of the following: ;, (, DATA, DEBUG, EPSILON, NOPRINT,
OUT.
ERROR 76-322: Syntax error, statement will be ignored.
36 MODEL BY SY = year treat rep / LACKFIT VIF TOL XPX COVB;
---
22
76
ERROR: No data set open to look up variables.
ERROR: No data set open to look up variables.
ERROR: No data set open to look up variables.
ERROR: No data set open to look up variables.
ERROR: No data set open to look up variables.
ERROR 22-322: Syntax error, expecting one of the following: ;, /, ACTUAL, BYOUT, COOKD, COVAR, D,
DUMMY, L95, L95M, LACKFIT, NOANOVA, NOAOV, NOCODE, NOOPT, NOOPTIMAL, NOPRINT, PREDICT,
PRESS, RESIDUAL, U95, U95M.
ERROR 76-322: Syntax error, statement will be ignored.
37 RIDGE MAX MIN;
38 run;
NOTE: The SAS System stopped processing this step because of errors.
NOTE: PROCEDURE RSREG used (Total process time):
real time 0.06 seconds
cpu time 0.04 seconds
39 PROC REG DATA=Work.Test2 outest=est1 outsscp=sscp1 rsquare;
40 Title'PP Plot'
41 MODEL BY SY = NLATOT WLATOT cycloxidim1/ noprint;
WARNING: The TITLE statement is ambiguous due to invalid options or unquoted text.
42 Plot npp. *r. / annotate=test2 nostat cframe=ligr
43 modellab="'Best'Two-Parameter Model:";
44 run;
WARNING: No variables specified for an SSCP matrix. Execution terminating.
NOTE: PROCEDURE REG used (Total process time):
real time 0.14 seconds
cpu time 0.14 seconds
NOTE: The data set WORK.EST1 has 0 observations and 8 variables.
NOTE: The data set WORK.SSCP1 has 0 observations and 3 variables.
45 Title'QQ Plot'
46 Plot r. *nqq. / noline mse cframe=ligr
WARNING: The TITLE statement is ambiguous due to invalid options or unquoted text.
47 modellab="'Best'Two-Parameter Model:";
48 run;
49 PROC Corr DATA=Work.Test2;
50 Var BY SY NLATOT WLATOT cycloxidim1;
51 run;
NOTE: PROCEDURE CORR used (Total process time):
real time 0.14 seconds
cpu time 0.07 seconds
Please help me to Know my fault in writing sas codes.
2. Also I want to find linear, quardratic and cubic relationship between BY and SY with NLATOT WLATOT Cycloxidim1 for Treatment 3 to 10.
thanks.
