#Remove objects from prior sessions
rm(list=ls())

#make sure the working directory is correct
getwd()
setwd("C:/Users/aamahama/Desktop/")

#Next load in the data set from a spreadsheet (.csv)
ds1<-read.csv("ds 1.csv",header=T)

#look at some of the data to make sure it is read into R properly
ds1[1:10,]
ds1[26:35,]

attach(ds1)
summary(ds1)

ds1$entry<-factor(ds1$entry)
ds1$env<-factor(ds1$env)
ds1$blk<-factor(ds1$blk)
summary(ds1)
detach(ds1)

mean(yld)
sd(yld)

# conduct some exploratory data analyses
# histogram of the yield data
hist(ds1$yld)
hist(ds1$yld,col="beige",main="Frequency of Yield Across Locations",xlab="Yield")

#Try a boxplots to look at distributions of each environment
boxplot(ds1$yld~ds1$env,main="Boxplot by Environment",xlab="Environment",ylab="Yield")
boxplot(ds1$yld~ds1$entry,main="Boxplot by Entry",xlab="Entry",ylab="Yield")
boxplot(ds1$yld~ds1$blk,main="Boxplot by Block",xlab="Block",ylab="Yield")

# consider analysis of variance: What is the model?
# linear model: y=mean+env+blk(env)+ent: What about GxE? 
anova(lm(yld~env+entry))
aov<-(anova(lm(yld~env+env/blk+entry)))

#generate lsmeans
library(lsmeans)
lsm<-(lm(yld~env+entry))
lsmeans(lsm,~entry)

sink("EDA _AOV_ds1.txt")
aov
lsmeans(lsm,~entry)
sink()