R - programming language and software environment for statistical

computation and visualization of results

Class 4. Charts, plots, graphs

# Operators for generating sequence of values

# Let’s remind operators seq() and rep()

# sequence with step 1:

1:12

# seq() generates a sequence, parameter 'by' specifies the step of the sequence
# sequence of numbers from 1 to 100 with step 2
seq(1,100,by=2)

# rep()repeats the number in the given sequence a certain number of times

# and then repeats this action until the desired length is reached

# numbers from 1 to 4, each repeated twice, until the desired length is reached
# in total 10 elements
rep(1:4, each = 2, len = 10)

# Operators sample() and rnorm()

# sample() creates a random permutation of the vector given as a parameter
sample(1:6)

# 10-times draw from a set of numbers 1:6 (e.g. 10-times dice roll simulation)
sample(1:6,10,replace=T)

# rnorm()generates a set of 30 numbers with an average of 50 and a standard deviation of 5

# with a normal distribution
rnorm(30,50,5)

# Charts
# In the R environment, you can use a large number of ready charts. Numerous parameters in graphic
# functions allow you to modify the appearance and to refine the visual presentation of data.

# first let’s install and load the ggplot2 package with the appropriate commands
install.packages("ggplot2")
library(ggplot2)

1
# Simple graph plot(x,y)
# as arguments of function plot (x, y) enter two vectors - the first gives x coordinates and second - y.
# if in plot() you will enter only one vector, the second by default will consist of successive
# natural numbers (x coordinates)

# Example 1. Create vectors x and y. Then draw a simple graph using the function plot().
x = c(7.5, 9, 7, 7, 10.5, 8.5, 10, 12.5, 9.5, 10.5)
y = c(65, 70, 66, 68, 66, 64, 71, 67, 72, 75)
plot(x,y)

# By default plot() is a dot plot. By adjusting the "type" parameter you can get other types:
# "l" - lines (lowercase letter L)
# "b" - both points and lines
# "c" - lines part without points
# "o" - both points and lines (overplotted)

# add lines connecting successive points by adjusting the parameter „type”

plot(x,y,type="l")

# obtained graph is very unclear, so we will sort the data according to the variable x.
# To do this, create a data frame, and then sort the data ascending with function order():
data = data.frame(x,y)

data = data[order(data[, 1]), ]

# create a dot plot - this one looks the same as the first plot
plot(data)

# create a plot with lines - this plot is more readable

plot(data, type = "l")

# create a plot with both points and lines

plot(data, type = "b")

# Editing the basic elements of the chart

# main - main title of the plot
# xlab – X axis label
# ylab – Y axis label
# col - graph color (you can set many colors by setting col as a vector or a color palette)
# col.main - color of a main title
# col.lab - color of labels of both x and y axes
# font – font type (1 - normal, 2 - bold, 3 - italic, 4 – bold italic)
# lty – line type (1 - line, 2 - dashed, 3 - dotted, 4 – dash-dot, 5 – long dash, 6 – double dash)
# lwd - line thickness (by default 1).

2
# Bar chart barplot()
# Example 1. Bar chart for weekly measurements of maximal temperature
max.temp = c(22, 27, 26, 24, 23, 26, 28)
barplot(max.temp)

# add to the chart: main title, axis titles, colors and bar descriptions, horizontal display
barplot(max.temp,
main = "Weekly maximal temperature",
xlab = "Celsius degrees",
ylab = "Day",
names.arg = c("Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"),
col = "darkred",
horiz = TRUE)

# Example 2. Bar chart of the age of a students’ group

age = c(21,18,19,21,22,19,22,20,23,22)
barplot(age)

# age vector summarized by count

table(age)

# bar chart of age vector summarized by count

barplot(table(age))

# add to the chart: main title, axis titles, colors and bar borders
barplot(table(age),
main="Age of 10 students - count",
xlab="Age",
ylab="Count",
border="red",
col="blue",
density=10)

# Example 3. Bar chart of weekly sales of various types of cars

cars = c(1, 3, 6, 4, 9)
trucks = c(2, 5, 4, 5, 12)
suvs = c(4, 4, 6, 6, 16)

# create data frame autos_data

autos_data = data.frame(cars, trucks, suvs)

# create a bar chart of the variable Suvs with specified labels for axes and orange bar borders
barplot(autos_data$suvs, main="Suvs sales", xlab="Days",
ylab="Total", names.arg= c("Mon","Tue","Wed","Thu","Fri"),
border="orange")

# create a bar chart of the variable Cars with labels for axes, blue borders and diagonal lines in bars
barplot(autos_data$cars, main="Cars sales", xlab="Days",
ylab="Total", names.arg=c("Mon","Tue","Wed","Thu","Fri"),
border="blue", density=c(10,20,30,40,50))

3
# create a bar chart of all the variables in autos_data with adjacent bars using rainbow colors
barplot(as.matrix(autos_data), main="Autos", ylab= "Total", beside=TRUE,
col=rainbow(5))

# Pie chart pie()

# They are considered as less efficient than bar charts, because pie charts require only one
# series of values (they do not allow to compare different data series divided by subgroups)

# Example 1. A pie chart of the age of students’ group

age = c(21,18,19,21,22,19,22,20,23,22)
pie(age)

# pie chart of age vector summarized by count

pie(table(age))

# add to the chart: colors in greyscale

pie(table(age), col=grey(0:4/4))

# Example 2. Pie chart of weekly sales of various types of cars

cars <- c(1, 3, 6, 4, 9)
pie(cars)

# add to the chart: main title, colors and labels

pie(cars, main="Cars sales", col=rainbow(length(cars)),
labels=c("Mon","Tue","Wed","Thu","Fri"))

# calculate the percentage for each day, rounded to one decimal place
car_labels <- round(cars/sum(cars) * 100, 1)

# concatenate a % sign after each value

car_labels <- paste(car_labels, "%", sep="")

# add to the chart: main title, colors and custom labels

pie(cars, main="Cars sales", col=rainbow(length(cars)), labels=car_labels,
cex=0.8)

# add to the chart: legend at the right side

legend(1.5, 0.5, c("Mon","Tue","Wed","Thu","Fri"), cex=0.8,
fill=rainbow(length(cars)))

# Example 3. A pie chart of the distribution of political parties in the parliament

parties = data.frame(Politicalparty = c("PO", "PSL", "PiS", "TR", "SLD",
"SP", "independent"), Members=c(190, 53, 190, 56, 86, 37, 28))

pie(parties[, 2], labels = parties[, 1], col = rainbow(length(parties[,

1])), main = "Distribution of political parties")

4
# Plotting a function in a given range curve()
# Example 1. Chart with a function log10(x) , in a range from 0.1 to 10:
curve(log10(x), 0.1, 10)

# chart with two (or more) functions in one graph window allows the parameter add
curve(sin(x), 0, 10, add=TRUE)

# By default, the function graph line is black. Change the line color to red and blue
curve(log10(x), 0.1, 10, col="red")
curve (sin(x), 0, 10, col="blue", add=T)

# Example 2. Chart with chi^2 distribution for different degrees of freedom

x = seq(from=0, to =10, by=0.1)
curve(dchisq(x,1), xlim=c(0,10), ylim=c(0,0.6)) #1 degree of freedom
curve(dchisq(x,4), add=T) # 4 degrees of freedom

# add to the chart: main title, colors and thick lines

curve(dchisq(x,1), xlim=c(0,10), ylim=c(0,0.6), col="red", lwd=3,
main="Chi^2 distributions", xlab="", ylab="")

curve(dchisq(x,4), add=T, col="green", lwd=3)

# add to the chart: legend at the top right side and colors
curve(dchisq(x,1), xlim=c(0,10), ylim=c(0,0.6), col="red", lwd=3,
main="Chi^2 distributions", xlab="", ylab="", legend("topright",c("1 degree
of freedom","4 degrees of freedom"),fill=c("red","green")))

curve(dchisq(x,4), add=T, col="green", lwd=3)

# Histogram hist()
# Example 1. Histogram and normal distribution
# sample: 200 numbers
proba = rnorm(200)

# create a histogram
hist(proba,20,probability=T, col="yellow")

# add to the chart: a normal distribution curve

curve(dnorm(x), lwd=3, col="red", add=T)

# Other graphs
# boxplot() - box chart
# stars - radar chart
# mosaicplot - mosaic chart
# pairs() - set of scatterplots

5
# Tasks

# Task 1. Define the cars vector with 5 values: 3, 5, 8, 6, 9.

# Plot cars with blue points overlayed by a line, and a title “Autos” with red color and bold italic font.

# Task 2. Create a vector W containing numbers from 4 to 13.

# Create a bar chart of the vector W with the following parameters:
# • enter the title of the chart "Chart",
# • enter "arguments" as the title of the x axis,
# • enter "values" as the title of the y axis,
# • set the color of the bars to red.

# Task 3. Create a vector wx with elements from 1 to 4 repeated eight times.

# For wx generate a pie chart and a pie chart of wx vector summarized by count.
# Use the rainbow() parameter to colorize the pie segments.
# Save the chart in .PNG format

# Task 4. Create a graph of a function y = cos(x) in a range (−2π, 2π) with the blue line.
# Then, in the same chart, add a graph of the function y = sin (x) drawn with a red line.
# Add the title of the chart and the titles of the X and Y axes.

Hint: xlim - range of x, therefore (−2π, 2π) you can write down as xlim=c(-2*pi,2*pi).

# Task 5. Create vector M as a set of 400 numbers with a normal distribution.

# For vector M create a histogram with the following parameters:
#• enter the title of the chart „Histogram”,
#• enter "Data x" as the title of the x axis,
#• enter "Data y" as the title of the y axis,
#• set the color to lightblue.

# Add to chart: a normal distribution curve of thickness 4 and red color.

# Save the chart in .PNG format

# Task 6. Create vector N as a set of 40 numbers with an average of 25 and a standard deviation of 5
# with a normal distribution.
# For vector N create a histogram with the following parameters:
#• enter the title of the chart „Histogram”,
#• enter "Data x" as the title of the x axis,
#• enter "Data y" as the title of the y axis,
#• set the color to green.

6
Task 1:

cars <- c(3, 5, 8, 6, 9)

plot(cars, type = "b", pch = 16, col = "blue", main = "Autos", font.main = 4, col.main = "red")

Task 2:

W <- 4:13
barplot(W, main = "Chart", xlab = "arguments", ylab = "values", col = "red")

Task 3:

wx <- rep(1:4, each = 8)

pie(wx, col = rainbow(length(unique(wx))))
png("pie_chart.png")
pie(table(wx), col = rainbow(length(unique(wx))))
dev.off()

Task 4:

x <- seq(-2*pi, 2*pi, length.out = 100)

y1 <- cos(x)
y2 <- sin(x)
plot(x, y1, type = "l", col = "blue", xlab = "x", ylab = "y", main = "Trigonometric Functions")
lines(x, y2, col = "red")

Task 5:

M <- rnorm(400)
hist(M, main = "Histogram", xlab = "Data x", ylab = "Data y", col = "lightblue")
curve(dnorm(x), lwd = 4, col = "red", add = TRUE)
png("histogram.png")
hist(M, main = "Histogram", xlab = "Data x", ylab = "Data y", col = "lightblue")
curve(dnorm(x), lwd = 4, col = "red", add = TRUE)
dev.off()

Task 6:

N <- rnorm(40, mean = 25, sd = 5)

hist(N, main = "Histogram", xlab = "Data x", ylab = "Data y", col = "green")