library(mosaicData)
data(SwimRecords)
View(SwimRecords)
help(SwimRecords)

Research Question: How have 100m records varied over the years?

Variables involved:

• year explanatory
• time response

Their types:

• year numerical
• time numerical

With numerical variables we have used:

• densityplot()
• bwplot()
• histogram()
• favstats()

Try this:

histogram(~time|year,data=SwimRecords)

• explanatory variable year is numerical, not a factor, and
• it has LOTS of values!

We need a new graphical method to study the relationship between two numerical variables.

Use the xyplot() function:

xyplot(time~year,data=SwimRecords,
       main="100m Swim Records",
       xlab="year",
       ylab="time (seconds)",
       pch=19)

Basic Form:

\[xyplot(response \sim explanatory, data = \ldots)\]

Rising cloud: the taller the father, the taller the son!

Falling cloud: bigger x's go with smaller y's!

There's a relationship, but it's not linear!

See the cloud "leveling off"?

As you move to the right, cloud neither rises nor falls.

data(stumps)
View(stumps)
help(stumps)

Research Question: Are there more larvae cluster in plots where there are more cottonwood stumps?

data(fuel)
View(fuel)
help(fuel)

Research Question: How does the speed at which the Ford Escort is driven affect its fuel efficiency?

Use a groups argument, with a key:

xyplot(time~year,data=SwimRecords,
       main="100m Swim Records",
       xlab="year",
       ylab="time (seconds)",
       pch=19,
       groups=sex,
       auto.key=TRUE)

data(TenMileRace)
View(TenMileRace)
help(TenMileRace)

Let's look at the relationship between age (explanatory) and net time (response), for both men and women.

xyplot(net~age,data=TenMileRace,
       groups=sex,
       auto.key=TRUE)

Make a scatter plot for each sex:

xyplot(net~age|sex,data=TenMileRace,
       main="Age and Race Time, by Sex",
       xlab="age (years)",
       ylab="10k time (seconds)",
       layout=c(2,1))

The guys appear to run faster. Hard to tell if age make a difference.

help(verlander)

• px horizontal distance of ball from center of plate (feet)
• pz vertical distance of ball above plate (feet)

Research Question: How are these variables related?

xyplot(pz~px,data=verlander,
       main="Verlander over the Plate",
       pch=19)

Over-plotting! (15,307 points sitting on top of each other.)

xyplot(pz~px,data=verlander,
       main="Verlander over the Plate",
       pch=19,
       alpha=0.10)

With alpha set to 0.10, you need 10 points stacked on each other to get full color!

Let's group by type of pitch:

xyplot(pz~px,data=verlander,
       main="Verlander over the Plate",
       pch=19,
       alpha=0.10,
       groups=pitch_type,
       auto.key=list(space="right"))

Hard to tell!

Let's plot in separate panels for each pitch type:

xyplot(pz~px|pitch_type,data=verlander,
       main="Verlander over the Plate",
       pch=19,
       alpha=0.10)