STATISTICAL COMPUTING ACTIVITY: CHAPTER 1 & 2

Step 1. Download husb.text from the web site.

Step 2. Run R

Step 3. Go to File menu and “Change directory” to the location that you have saved the file

>husb=read.table(“husb.txt”, header=T)

>attach(husb)

To see the variable names and data type

>husb

Let us get some descriptive statistics

>mean(husb)

>var(husb)

>sd(husb)

>median(husbage)

(Note that you need variable name not the data file name)

>fivenum(wifeage)

(Note that you need variable name not the data file name)

>summary(husb)

>cov(husb)

>cor(husb)

This will produce Figure 2.1 (a) which is a simple scatterplot of wife’s age versus husband’s age.

>plot(husbage,wifeage)

Note how the arguments are entered. The form is plot(x,y), first the explanatory variable (x), then the response variable (y).

Now, let us add an y=x line, i.e. any point on this line will imply husband’s age=wife’s age. This is the Figure 2.1 (b)

>abline(0,1)

Here the form is abline(intercept,slope).

If you would like to add a least squares regression line you need to use

>abline(lm(wifeage~husbage))

The symbol “~” means relationship, “lm" means linear model.

Next let us add noise to husbage and wifeage to prevent overlapping. This is called jittering.

> plot(jitter(husbage),jitter(wifeage))

This is the Figure 2.1 (c).

Let us show the marginal distribution of the variables

>rug(jitter(husbage), side=1)

>rug(jitter(wifeage), side=2)

In this case side 1 is the horizontal axis, side 2 is the vertical.

Now, we have the Figure 2.1 (d).

To analyze the age difference between wife and husband, let us find the age difference

>agediff=husbage-wifeage

Let plot husband’s age at the first marriage versus the age difference

>plot(agediff, husagefi)

Let us add a line on which there is no age difference, i.e., agediff=0

> abline(v=0)

Here “v” is for vertical “h” for horizontal.

Now, we have the Figure 2.2.

To produced Figure 2.3 (a) and (b):

>plot(husbhei, wifehei)

>abline(0,1)

For Figure 2.4

>heidiff=husbhei-wifehei

>plot(agediff,heidiff)

>abline(v=0)

>abline(h=0)

Now, tell me what is the main thing that you did not like in this example. (My brain was fighting with my hands when I was preparing this handout.)

Labeling the Scatterplot with Row Names

Enter the following data to Excel

species	bodywt	brainwt
PotarMonkey	10.0	115
Gorilla	207.0	406
Human	62.0	1320
RhesusMonkey	6.8	179
Chip	52.2	440

Save as primates.txt

To put labels to the points on the scatterplot

Step 1. Run R

Step 2. Go to File menu and “Change directory” to the location that you have saved the file

>primates=read.table(“primates.txt”, header=T)

>attach(primates)

>row.names(primates)=species

>plot(bodywt, brainwt)

>text(x=bodywt, y=brainwt, labels=row.names(primates))

From Bivariate to Multivariate

To produce a scatterplot matrix you can either use

>pairs(husb)

Or

>plot(husb)

To smooth the data

> plot(husb, lower.panel=panel.smooth)

Coplots

To produce a conditional scatterplot of wifeage versus husbage given the husbhei use

>coplot(wifeage~husbage|husbhei)

To smooth the coplot by using lowess

> coplot(wifeage~husbage|husbhei, panel=panel.smooth)

Quantile-Quantile Plots (Normal Plots)

To produce a normal plot use

>qqnorm(husbage)

To add a line

> qqline(husbage)