[R-sig-ME] Help: Split-Split model nested in a nested structure

ian m s white i.m.s.white at ed.ac.uk
Sat Feb 22 18:09:06 CET 2014


For one district, the one way to obtain the split-split anova is

village <- gl(2,48)
field <- gl(16,6)
sow <- gl(2,24,96)
foo <- data.frame(village,field,sow)
foo$nit <- gl(2,3,96)
foo$geno <- gl(3,1,96)
foo$Y <- rnorm(96)
fit <- aov(Y ~ village*sow*nit*gen + Error(field/nit), data = foo)
summary(fit)

Note that fields are numbered 1 to 16 over the two villages.
 
On 22 Feb 2014, at 08:30, Zia Ahmed <zia207 at gmail.com> wrote:

> Dear list members.
> We are trying see effect of N on yield of three wheat genotypes under late
> and early planting conditions in saline and non-saline environments. Our
> experimental structure as follows:
> 
> Districts (2: DIST01 and DIST02) - Not randomly selected
> 
> ENV (Saline and Non-Saline) - One  saline and non-saline environment (eg.
> village) were selected in each district (not random).
> 
> *SOWING Time  (Late and Early):* 8 farmers' fields were selected randomly
> in each environment or village.  These are spatially distributed in a
> village. Among them, 4 were grouped as early and 4 were grouped as a late
> sowing group.
> 
> *N_TREAT  (N0 and N100)* - Then, each farmer field was splitted  into
> two  main
> plots where two N-treatment were assigned randomly.
> 
> *GENOTYPE ( G1,  G2, G3*) - Then, N-main plots were further splited into
> three subplots here 3 wheat varieties were  assigned  randomly.
> 
> This experiment was repeated for two years. Farmers' fields in Year 1 and
> Year 2 were not same. We are not interested to see the District effect on
> wheat yield here. Rather, we like to see the effect of saline environment,
> sowing time, N  and genotype and their interaction  on wheat yield for year
> 1 and year 2 separately.
> We are using the following ANOVA model. We do not know whether we are
> missing something here.  Help will be highly arreciated.
> Regards
> Zia Ahmed, CIMMYT
> 
> model<-aov(YIELD~ENV*SOWING*N_RATE*GEN+Error(FARMERS/N_RATE/GEN),
> data=mydata)
> 
> summary(model)
> 
> 
> 
> # District: two
> 
> DIST<-as.factor(rep(c("DIST01","DIST02"),each=96))
> 
> # ENV: Saline and Non-saline environment
> 
> ENV<-as.factor(rep(rep(c("Saline","Non-saline"),each=48),2))
> 
> # Farmers 16
> 
> FARMERS<-as.factor(rep(c("F1","F2","F3","F4","F5","F6","F7","F8",
> 
>                         "F9","F10","F11",
> "F12","F13","F14","F15","F16"),each=12))
> 
> # Showing Date: two
> 
> SOWING<-as.factor(rep(rep(c("Early","Late"),each=6),16))
> 
> # Nitrogen treatments: N0 and N100
> 
> N_RATE<-as.factor(rep(rep(c("N0","N100"),each=3),32))
> 
> # Genotype
> 
> GEN<-as.factor(rep(rep(c("V1","V2","V3"),each=1),64))
> 
> # Response: Wheat Yield
> 
> set.seed(1234)
> 
> YIELD <- rnorm(n=192, mean=3.0, sd=0.5)
> 
> # Create Data Frame
> 
> mydata<-data.frame(DIST,ENV, FARMERS,SOWING,N_RATE,GEN,YIELD)
> 
> mydata
> 
>> 
>> 
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
> 


-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.



More information about the R-sig-mixed-models mailing list