RStudio Cheat Sheet
by Adela Vrtkova and Martina Litschmannova, Department of Applied Mathematics, FEECS, VB-TUO
via cheatography.com
Workspace, Using libraries
?boxplot
getting help documentation for function
boxplot
getwd()
returning the current working directory
setwd("C:/Users/RStudio")
setting the working directory to specied le
install.packages("packageZ")
downloading and installing a package called
packageZ
library(packageZ)
activating already installed package called
packageZ
packageZ::functionF(x)
calling function
functionF
from specied package
packageZ
moments
,
EnvStats
,
dunn.test
,
lsr
,
openxlsx
,
car
,
epiR
important packages
# After the hash, I can write whatever.
writing notes into the script
Importing data
data = read.csv2("C:/Users/RStudio/data.csv")
importing data in
csv
from specied le and saving as
data
data = read.csv2("http://am-nas.vsb.cz/DATA/dataset.csv")
importing data in
csv
from the internet and saving as
data
data = readWorkbook("C:/USER/DATA/dataset.xlsx", sheet=1,
startRow=4, colNames=TRUE, cols=2:9) # openxlsx package
importing data in
xlsx
Working with data
data = as.data.frame(data)
saving imported data as an object of class
data.frame
data.S = stack(data)
transferring data table into the standard data matrix
data.S.omit = na.omit(data.S)
omitting entire rows with missing values (NAs)
Probability distribution - Prexes
r- generating random numbers from the distribution
d- probability density function
f(x)
or probability mass function
P(X=x)
p-
P(X≤x)
q- quantile function
Probability distribution - Discrete
-binom Binomial distribution
Bi(n, π)
-hyper Hypergeometric distribution
H(N, M, n)
! R code requires -
H(M, N −M, n)
-nbinom Negative binomial distribution
NB (k, π)
! denition in JASP/R - number of unsuccessful trials
-pois Poission distribution
P o(λt)
Probability distribution - Continuous
-unif Uniform distribution
U(a, b)
-exp Exponential distribution
Exp(λ)
-norm Normal distribution
N(µ, σ2)
! JASP applet Distributions requires
N(µ, σ2)
! R code requires -
N(µ, σ)
EDA for a Qualitative Variable
data$group = as.factor(data$group)
redening group variable as
factor
table(data$group)
frequency table
barplot(table(data$group))
creating a bar plot
pie(table(data$group))
creating a pie chart
EDA for a Quantitative Variable
summary(data$values) summary statistics
length(data$values) sample size (attention if
NAs present)
min(data$values) minimum
mean(data$values) arithmetic mean
quantile(data$values,probs=0.3) 30% quantile
max(data$values) maximum
sd(data$values) standard deviation
var(data$values) variance
moments::skewness(data$values) skewness
moments::kurtosis(data$values)-3 kurtosis
boxplot(data$values) boxplot
hist(data$values) histogram
plot(density(data$values)) plotting kernel density
estimation
qqnorm(data$values); qqline(data$values) QQ-plot
Function tapply()
tapply(dataS$values, dataS$group, mean)
calculates the mean for
values
by
group
in
data
tapply(dataS$values, dataS$group, quantile, probs=0.4)
calculates the 40% quantile for
values
by
group
in
data
tapply(dataS$values, dataS$group, moments::kurtosis)-3
calculates the kurtosis for
values
by
group
in
data
Statistical inference - One variable
shapiro.test(data$values)
Shapiro-Wilk test
varTest(data$values, sigma.squared=400, alternative="two.sided",
conf.level=0.95) # EnvStats package
condence interval for variance and one-sample Chi-squared test on
variance
(H0∶σ2=400, HA∶σ2≠400)
t.test(data$values, mu=5, alternative="less", conf.level=0.95)
condence interval for mean and one-sample Student's t-test
(H0∶µ=5, HA∶µ<5)
wilcox.test(data$values, mu=8, alternative="greater", conf.level=0.95,
conf.int=TRUE)
condence interval for median and one-sample Wilcoxon test
(H0∶x0,5=8, HA∶x0,5>8)
binom.test(x,n,p=0.18,alternative="two.sided",conf.level=0.95)
condence interval for probability and one-sample Binomial test
(Clooper-Pearson method)
(H0∶π=0.18, HA∶π≠0.18)
