G Additional Information on Graphing

Bar Charts (2.1)

Useful for categorical variables “variable” in the statement below is the variable we want to plot. We need, first to put it in a table so R can draw the plot. “myvariable1” is just the name I chose for the table.

myvariable1 <- table(variable)

barplot(myvariable1)

To get relative frequency (instead of just a count) use

table(variable)/number

where number is the total number of observations myproportion <- table(variable)/number

barplot(myproportion)

main, xlab, ylab, las can all be used as was shown in the previous crib sheet

names.arg = c(what names you want) can also be added if you want to change the default names of the variables.

Pie chart (2.1)

Pie charts are not often used in statistical analysis but are common in news reports

first, create a table:

variabletab <- table(variable)

then plot the pie chart of variabletab

pie(variabletab)

will produce a pie chart of the variable named. You can add a box around the plot with

box()

Boxplots 2.2 and 2.3

Boxplots are used for numeric variables.

boxplot(numeric variable)

quantile(numeric variable, probs=c(0, 0,25, 0.5, 0.75, 1)

will return the values for the components of the plot (the upper and lower limits of the box as well as the upper and lower limits of the whiskers).

main, xlab, ylab, las, can all be used as was shown previously

ylim will reset the limits for the y variable:

ylim(lower limit, upper limit)

boxplot(Var1~Var2)

will produce a box plot of Var1 broken down by Var2

boxplot(Var1[Var2==“One thing”], Var1(Var2==“The other thing”]

will do the same thing.

Histograms 2.4

Histograms are used for numeric variables

hist(Variable)

hist(Variable, freq = F)

will give relative frequency as will

hist(Variable, prob = T)

ylim(startvalue, endvalue)

to set y limits

breaks=N

for number of break points or bins

breaks=c(a,b,c,…n)

will break at specific points

breaks=seq(from=begin number, to=end number, by= increment)

another way

lines(density(Variable)

adds a density plot. NOTE: Y axis must be Density, not Frequency.

col, lwd, main, xlab, ylab can all be added also.

Stem and Leaf (2.5)

Stem and Leaf plots are used for numeric variables. (I prefer to create them by hand, as they are best used for relatively small data sets and getting R to do them just the way you want can be difficult.)

stem(variable, scale = number)

###Bar Charts and Mosaic Plots (2.6)

Bar charts and mosiac plots are used when you have 2 categorical variables

First, you need to create a table for the two variables

Table 1<-table(Var1,Var2)

barplot(Table1)

produces a stacked barplot (this is the default)

barplot(Table1, beside = T)

produces side by side bar plots

barplot(Table1, beside = T, legend.text = T)

for default legend

barplot(Table1, beside = T, legend.text =c(“Name1”,”Name2”)

produces a custom legend

main, xlab, ylab, and las may also be added to the command with the usual results

col may also be added for color

mosaicplot(Table1)

produces a mosaic plot

all of the above options may also be added.

Scatterplots (2.7)

Scatterplots are used when you have two quantitative variables

plot(Var1, Var2)

var1 on x axis, var2 on Y

xlab, ylab, and main may be added

range of the variables may be altered with xlim and ylim (FORM: xlim=c(lower.value, upper.value)

size of plotting character can be changed with cex argument

plotting character can be changed with pch argument

col can be added

abline(lm(var1, var2))

will add the regression line

col may be specified for the line color

lty and lwd for line type and line width, may also be specified