objects

declaration
vectorname<-c(1,2,3,4)
typing the object name outputs a list of the data set
mean(mydatainR$weight)
mydatainR is the object
weight is a variable inside the object
mean(weight) only if the object is attached already

`mean(mydatainR[,1])
mean of the first column in the data set
`sd(mydatainR$weight[1:3])
sd of just the first 3 weights
r is case sensitive
commands
data management
objects()
R lists all saved objects
head(object)
shows the first few rows of the data
text()
text(x, y = NULL, labels = seq_along(x$x), adj = NULL, pos = NULL, offset = 0.5, vfont = NULL, cex = 1, col = NULL, font = NULL, ...)
split()
split(x, f, drop = FALSE)
x: represents data vector or data frame
f: represents factor to divide the data
drop: represents logical value which indicates if levels that do not occur should be dropped
split(puffinsA1, puffinsA1$Cloud)
attach(mydatainR)
R will look in this date set for variable names
can then use commands directly with the variables
problematic if its the same name as another object
formulas
pnorm(value,mean,sd)
area to the left of value of the PDF
default = standard normal distribution, mean = 0, sd = 1
qnorm(area,mean,sd)
outputs value that has the area given to the left of it
mean(vectorname)
mean of the vector
choose(n,x)
combination formula/binomial coefficient
$$ \begin{pmatrix} n\ x \end{pmatrix}= \frac{n!}{x!(n-x)!}$$
dbinom(x,n,P(X))
pbinom(x,n,P(X))
for when we want probability of <= x
var(vectorname)
gives the variance of the vector
sd(vectorname)
standard deviation of the vector
same thing as sqrt(var(vectorname))
summary(vectorname)
yields the mean and the 5-number summary (minimum, first quartile, median, third quartile, maximum)
max, min, sum, median
punif(value,parameter1,parameter2)
probability of uniform distribution
$P(X\leq value)$ between parameter1 and 2
qunif(value,parameter1,parameter2)
inverse function of punif, gives value when inputting the area
defining functions
fib_closed <- function(n) { 1/sqrt(5) * (((1 + sqrt(5))/2)^n - ((1 - sqrt(5))/2)^n)}
print(fib_closed(seq(30)))
seq(30) = [1,30
importing csv files
mydatainR<-read.csv(file.choose(),sep=",",header=TRUE)
sep=","because its comma separated values(csv)
can be omitted if the file is .csv
header=TRUE means that the first row represents column names, not data values
both are the default so can just do mydatainR<-read.csv(file.choose())
empty choose opens file explorer, can also specify the path
mydatainR<-read.csv("C:/Documents and Settings/JB/My Documents/mydata.csv",header=TRUE)
help(commandname) / ?commandname
gives syntax, possibles options and examples for the command
 vectors can be used in commands
`mysecondvector<- 10+2*myfirstvector
plots
commands
hist(mydatainR$weight)
hist(change,xlab="Percent change in body weight",ylab="Frequency", main="Histogram of Percent Change in Body Weight at Autopsy",
cex.lab=1.5,nclass=20,col="green")
xlab = adds x-axis label
ylab = adds y-axis label
main = adds title
nclass = changes the number of classes
col = changes colour
cex.main/lab/axis controls the size of the title/labels
default cex = 1, ratio

boxplot(mydatainR$weight)
`boxplot(mydatainR$weight~mydatainR$group)
side-by-side boxplots of the weight variable for each group individually
one for weight values in group A, other for the values in group B
boxplot(Butterfat~Breed,ylab="BUTTERFAT, YO", names=c("Breed 1", "Breed sqrt(4)", "Third one","Last one?","No, really, last one"),
col="pink",cex.axis=1.5,main="Not a great title but it’s not bad",cex.main=2)
xlab = adds x-axis label
ylab = adds y-axis label
main = adds title
names = adds groups names
col = changes colour
cex.main/lab/axis controls the size of the title/labels
default cex = 1, ratio
 plot(height,weight)
basic scatterplot of weight on height
plot(Moonlight,Count,xlab="Moonlight%",ylab="Number of Passes",cex.lab=1.6,cex=2,col="blue",cex.axis=1.5)
xlab = adds x-axis label
ylab = adds y-axis label
main = adds title
pch = controls the plotting symbol
help(pch) for more
pch = 1 -> open circle, 15 = filled-in square, 16 = fille-in circle
col = changes colour
cex.main/lab/axis controls the size of the title/labels
default cex = 1, ratio

qqnorm(mydatainR$weight)
a normal quantile-quantile plot of the weight variable
qqline(mydatainR$weight)
adds a line to the normal qq plot already created
barplot(object)
barplot(count,names=c("Ugh","Meh","Not too shabby"),xlab="Daily Vibe",ylab="Frequency", cex.names=1.2,cex.lab=1.5,col="aquamarine")

calculating
binomial probabilities
dbinom(x,n,p)
yields P (X = x) for a binomial distribution
dbinom(3,10,.5) yields 0.1171875
gives the probability that the random variable is exactly equal to x
pbinom(x,n,p)
yields P (X ≤ x) for a binomial distribution
pbinom(3,10,.5) yields 0.171875
gives the probability that the random variable is less than or equal to x
the cumulative distribution function (??)
 poisson probabilities
dpois(x,lambda)
yields P (X = x) for a Poisson distribution
dpois(3,5) yields 0.1403739
ppois(x,lambda)
yields P (X ≤ x) for a Poisson distribution
ppois(3,5) yields 0.2650259
normal and t distribution
pnorm(z)
yields the area to the left of z under the standard normal curve
pnorm(-1.96) yields the value 0.02499790, the area to the left of −1.96 under the standard normal curve
pnorm(x,mu,sigma)
yields the area to the left of x under a normal curve with μ = mu and σ = sigma
pnorm(2.5,2,1.8) yields 0.6094085, which is the area to the left of 2.5 under a normal curve with a mean of 2 and standard
deviation of 1.8
mu = mean
sigma = standard deviation
qnorm(p)
gives the value of z such that the area to the left of z under a standard normal curve is p
inverse function of pnorm
qnorm(p, mu, sigma)
yields the value of x such that the area to the left of x under a normal curve with a mean of mu and standard deviation of sigma
qnorm(.43,12,24) would yield the 43rd percentile of a normal distribution with a mean of 12 and a standard deviation of 24
dnorm(x, mu, sigma)
yields the probability density function (height of the curve at point x) for a normal distribution with a mean of mu and standard deviation
of sigma
normally used when plotting the normal distribution pdf
rnorm(n,mu,sigma)
randomly generates n values from a normal distribution with a mean of mu and a standard deviation of sigma
often used in (my?) simulation functions
pt(x,df)
yields the area to the left of x under a t distribution with df degrees of freedom
pt(-1.96,5) yields the value 0.05364398, the area to the left of −1.96 under a t distribution with 5 degrees of freedom
qt(p,df)
inverse of pt
yields the value x such that the area to the left of x under a t distribution with df degrees of freedom is p
qt(0.975,23) yields 2.068658
t procedures
one sample
t.test(mydatainR$weight,mu=10)
tests the null hypothesis that the true mean of the weight variable is 10
It also outputs a 95% confidence interval for the true mean of the weight variable
The default is a two-sided alternative hypothesis, and the default confidence level is 95%.
two sample
t.test(weight~group)
tests the null hypothesis that μA = μB and outputs a 95% confidence interval for μA − μB
R chooses the name that comes first alphabetically to be Group 1 for the grouping variable
default = Welch method (unpooled variance)
= t.test(weight~group,var.equal=FALSE)
does not assume the population variances are equal
t.test(weight~group,var.equal=TRUE
uses the pooled-variance procedure
assumes the population variances are equal