DATA1001 Codes

sqrt(x)

square root

abs(x)

absolute value of x

length(x)

the amount of numbers

min(x)

the lowest value of the numbers

sum(x)

the sum of the numbers

str(dataset_name)

structure of the data- shows the number of rows and columns in the data, the variable names, and R's classification of each variable

head(dataset_name, n)

shows the first n rows of the data set (shows 6 if n is not specified)

tail(dataset_name, n)

shows the last n rows of the data set (shows 6 if n is not specified)

ggplot()

creates graphical summaries

library(tidyverse)

loads the tidyverse library

geom_histogram()

histogram graph

geom_boxplot()

boxplot graph

ggplot(iris, aes(x = Petal.Width, fill = Species)) + geom_histogram() + labs(title = "Sliced histogram of Petal Width by Species")

Iris histogram example

mean(x)

mean

median(x)

median

sd(x)

sample standard deviation

 library(rafalib), popsd(dataset_name$variable)

population standard deviation

fivenum(x)

minimum, Q1, median, Q3, maximum

IQR(x)

IQR

 LT = Q1 - 1.5 * iqr_value

Lower threshold outlier

UT = Q3 + 1.5 * iqr_value

upper threshold outlier

as.factor(), e.g. mtcars$vs = as.factor(mtcars$vs)

class(mtcars$vs)

converting numeric to a factor

char_list = c("1", "2", "3", "4")

character list

?dataset_name

help page for the data set

class(dataset_name$variable)

to see how r has classified a certain variable

pnorm(x=…, mean=.., sd=…, lower.tail=…)

works out the probability/the area under the curve up to a certain point/ the Normal cumulative distribution function (cdf). Defaults are mean=0, sd = 1, lower.tail=TRUE

 round(number, n)

round an answer to n decimal places

qnorm(x=…, mean=…, sd=…, lower.tail=…)

gives you the value (quantile) below which a certain percentage of data from a Normal distribution falls. Defaults are mean=0, sd = 1, lower.tail=TRUE

filter()

filtering data

mutate()

add new columns or modify existing ones

cor(x,y)

finds the correlation coefficient

lm(y ~ x, data=df)

creates a linear regression model where y is the dependent variable, x is the independent variable and df is the data.frame

geom_smooth(method="lm", se = T/F, color=…)

adding a regression line to a scatter plot where se is standard error (always put false) and method="lm" means we want a linear regression line in particular

geom_point()

makes a residual plot

geom_hline(yintercept=0)

creates a horizontal line at 0

ggplot(model, aes(x=.fitted, y=.resid)) + geom_point() + geom_hline(yintercept=0, linetype= "dotted", color="red")

scatter residual plot

sample(x, size, replace = T/F, prob=)

modelling random events where x is the data from which you sample, size is the size of your sample, replace is with or without replacement, prob (optional) is the set probability for each chance

replicate(x, function)

to repeat a function multiple times where x is the number of repeats

set.seed(x)

makes results reproducible where x is any number (question will tell you what to use)

dbinom(x,n,p)

 calculates P(X=x) where x is the number of events we want, n is the number of trials and p is the probability of success

pbinom(x,n,p)

combines many dbinom() to calculate P(X is less than or equal to x) where x is the number of events we want, n is the number of trials and p is the probability of success

Box=c(1,0)

to define your box

• Set.seed(1)

• Box=c(1,0)

• Sum(sample(box, 10000, replace=TRUE)

to simulate 1000 tosses

cumsum()

calculates the cumulative sum

rep(a,b)

to create large boxes where a is the number to be repeated and b is how many times it is repeated