#R-help to exercise 3.1 in BSS

# Read the data into a dataframe and give names to the variables:

names(cafe)<-c("no","sale")

# Take a look at the data (make sure they correspond to those given in the exercise):

cafe

# Attach the dataframe (making the variables available):

attach(cafe)

# Make a plot of sale as a function of the number or dispensers:

plot(no,sale)

# Inspect the plot. How is the relation between the number of dispensers and the coffee sale?

# Fit a straight line and draw it on the plot:

linfit<-lm(sale~no)

linfit

abline(linfit)

# How well does the straight line describe the relation between the number of dispensers and the coffee sale?

# Fit a second order polynomial (note that inside lm-command, we have to write the second order term inside I( ), # otherwise the sign ^  will be misinterpreted by R):

lm(sale~no+I(no^2))

# Compute and draw the fitted second order polynomial:

x<-seq(0,7,0.1)

koef<-lm(sale~no+I(no^2))\$coef

koef

lines(x,koef[1]+koef[2]*x+koef[3]*x^2,lty=2)

# Do the straight line or the second order polynomial provide the best description of the relation between the number of dispensers and the coffee sale?