# A short introduction to R
# The aim of this note is to give a brief introduction to R. You may copy the commands below from the web-browser and paste them into the command window of R.
# Everything on a line that comes after # is a comment and R disregards this.
# R as a calculator
# You may use R as a calculator. For example:
3+2
3-2
3*2
3/2
3^2
sqrt(2)
exp(2)
log(2)
#Scalars
# You may define scalar variables and use them in computations. For example:
a = 2 # alternatively a<-2
b = 3 # alternatively b<-3
a+b
a*b
a^b
#Vectors
# You may define vector variables and use them in computations. For example:
x = c(1,2,3,4)
y =c(2,4,6,8)
x+y
y-x
y/x
y^x
# Note that R does the computations elementwise
# (R may also perform vector and matrix algebra)
# You may select one or more element(s) of a vector:
x[2]
y[c(1,3)]
#Sequences
# R may create special sequences (stored as vectors):
cc = 1:10
cc
dd = seq(0,20,2)
dd
ee = rep(1,10)
ee
#Functions of a vector
# R has a number of functions that operate on vectors. For example:
sum(x)
prod(x)
length(x)
#Descriptive statistics
# Make a vector with the data from the lectures on the age of mineral samples:
rock.age=c(249,254,243,268,253,269,287,241,273,306,303,280,260,256,278,344,304,283,310)
# Compute mean, median, and standard deviation:
mean(rock.age)
median(rock.age)
sd(rock.age)
# The command "summary" gives a summary:
summary(rock.age)
#Reading data from files and dataframes
# We may read data from a file (which may be on the web) into a dataframe. For example:
sigarett=read.table("http://www.uio.no/studier/emner/matnat/math/STK4900/v12/sigarett.dat", header=T)
#Look at the data and output a summary of them:
sigarett
summary(sigarett)
#You may access one of the variables in a dataframe. For example:
sigarett$nicot
#(note that it is not sufficient to just write "nicot")
#You may attach the dataframe:
attach(sigarett)
#Then it suffices to just write "nicot"
#Some plots
#R may produce a number of useful plots. Some examples:
hist(nicot) # histogram
boxplot(nicot) # boxplot
qqnorm(nicot) # normal probability plot
#Multiple plots in one figure:
par(mfrow=c(1,2)) # two plots side by side
plot(co, nicot) # scatter plot
plot(tar, nicot)
par(mfrow=c(1,1)) # restore the setting with single plots
# Some statistical methods
# t-tests and confidence intervals:
t.test(rock.age)
# Linear regression:
fit=lm(nicot~co+tar)
summary(fit)
# Note that the summary-command depends on the type of object it is applied to.
# This is typical for the way R operates.
# Bootstrapping
# Bootstrap confidence interval for mean
bootagemean<-numeric(0)
for (i in 1:1000) bootagemean[i]<-mean(sample(rock.age,replace=T))
sort(bootagemean)[c(25,975)]
# Bootstrap confidence interval for median
bootagemedian<-numeric(0)
for (i in 1:1000) bootagemedian[i]<-median(sample(rock.age,replace=T))
sort(bootagemedian)[c(25,975)]
# Help functions
# R has a well developed help system that describes the commands. For example:
help(lm)
# If you do not remember the name of a command one may use the "help.search" command.
# For example for linear regression:
help.search("linear regression")
# Logging off and saving the workspace
# You exit R by giving the command:
q()
#You then get the question "Save workspace image?".
#If you answer yes, R will store all your variables so that you may resume the next session where the last one ended.
# Configuring R
# If you use R for more than one project, it is useful to create one folder for each project.
# For STK4900, you make a folder called STK4900.
# In this folder you paste a shortcut of the R icon.
# You then right-click on the icon, and write (e.g.) "C:\Documents and Settings\username\My Documents\STK4900" in the "Start in" field.
# This will ensure that all computations for STK4900 will be stored in the folder STK 4900, and hence kept apart from other R-computations.
# Libraries
# A number of libraries (or packages) are available in R.
# Some of these come with the standard implementation of R.
# One such example is the survival-package that may be loaded by:
library(survival)
# Other packages may be easily downloaded.
# To download a package, you chose "Install package(s)" from the Packages-menu in R.
# Then you chose a CRAN mirror from a drop-down menu (e.g Norway).
# Finally you chose the package from the next drop-down menu.