# Exercise 0116

# (a) read in the data
# the argument `header=T' tells R that the first row contains variable names.
ex0116 <- read.table("ex0116.txt",header=T)

# for comma separated variable files
# the option `sep=","' tells R that a comma is used to separate fields.
# ex0116 <- read.table("ex0116.txt",header=T,sep=",")

# check to see that the data was read in correctly
ex0116

# allow the variables in ex0116 to be accessed by name rather than by the long form
# (for example, ex0116$Planet).
attach(ex0116)

# (b) Draw a scatterplot of distance versus order
plot(Order,Distance)

# (c) repeat using logs for distance
plot(Order,log(Distance))

# (d) find mean and SD of the distances
mean(Distance)
sd(Distance)

# (e) find mean and SD of the log distances
mean(log(Distance))
sd(log(Distance))

# Exercise 0218

# verify computations in Display 2.9

case0201 <- read.table("case0201.txt",header=T)
case0201
attach(case0201)

# means and standard deviations
# the square brackets find subsets according to a criteria
mean(Humerus[Status=="PERISHED"])
mean(Humerus[Status=="SURVIVED"])
sd(Humerus[Status=="PERISHED"])
sd(Humerus[Status=="SURVIVED"])

# means and SDs again, but using split and sapply
# split divides a variable according to the categories of another variable (creating a list)
# lapply applies a function to a list
sapply(split(Humerus,Status),mean)
sapply(split(Humerus,Status),sd)

# difference in means
# this uses unlist (with the use.names option = F) to take the `unlist' produced by lapply
# diff takes the difference between adjacent entries in a vector.
diff(unlist(sapply(split(Humerus, Status), mean),use.names=F))

# find quantiles of a t distribution
qt(0.975,57)

# carry out a t test and find a 95% confidence interval
t.test(Humerus~Status,var.equal=T)

# compute an area under a t curve (p-value)
# pt(x,df) gives the area to the left of x under a t curve with df degrees of freedom

# one-sided example
1-pt(1.778,57)

#two-sided example
2*(1-pt(1.778,57))