Scribd
Upload
31 views

Summary of R Commands For Statistics 100

This document summarizes R commands that will be used in Statistics 100, including commands for reading and manipulating data, descriptive statistics, graphics, probability distributions, random sampling, statistical inference, and regression. Some key commands are read.table() for importing data, summary() for data summaries, plot() for scatter plots, t.test() for t-tests, lm() for linear regression and glm() for logistic regression. Help pages for each command can be accessed with help(command).

Uploaded by

Simon Maya
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
31 views

Summary of R Commands For Statistics 100

This document summarizes R commands that will be used in Statistics 100, including commands for reading and manipulating data, descriptive statistics, graphics, probability distributions, random sampling, statistical inference, and regression. Some key commands are read.table() for importing data, summary() for data summaries, plot() for scatter plots, t.test() for t-tests, lm() for linear regression and glm() for logistic regression. Help pages for each command can be accessed with help(command).

Uploaded by

Simon Maya
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Summary of R commands for Statistics 100

Statistics 100 – Fall 2011


Professor Mark E. Glickman

The following is a summary of R commands we will be using throughout Statistics 100, and maybe
a few extras we will not end up using. Please refer to the homework and course notes for examples
of their usage, including the appropriate arguments of the commands. In the descriptions below,
fnc is an arbitrary R command.

Reading, viewing, and assigning data in R:

y = fnc(x) − assigns the results of the function fnc evaluated at x to the variable y.

file.choose() − navigates to a data file on your computer.

read.table(fname) − reads data into R from file fname.

read.csv(fname) − reads data into R from a comma-separated value file fname

data.frame(...) − creates a data frame within R.

View(x) − view data frame x within R. Can also just type the name of the data frame at the
prompt.

help(fnc) − help page for function “fnc”.

Descriptive statistics:

summary(x) − data summary of x.

mean(x) − sample mean of x.

sd(x) − sample standard deviation of x.

length(x) − number of values in x.

table(x) − for categorical variable x, creates vector of counts of each unique category.

cor(x,y) − correlation between x and y.

by(y,x,fnc) − with categorical x and function fnc, carry out fnc(y) for each level of x.

1
Graphics:

hist(x) − histogram of data in x.

stem(x) − stem and leaf plot of data in x.

plot(x,y) − scatter plot of y against x.

lines(supsmu(x,y)) − add smoother to existing scatter plot.

boxplot(list(x1,x2,...)) − side-by-side boxplots of variables x1, x2, etc.

boxplot(y ~ x) − alternative method for boxplots if y is quantitative and x is categorical.

barplot(x) − barplot of x (where x contains the heights of the bars).

abline(a,b) − add the line y = a + bx to an existing plot.

abline(h=a) − add a horizontal line at y = a to an existing plot.

abline(v=a) − add a vertical line at x = a to an existing plot.

abline(model.fit) − add a regression line based on the model model.fit to an existing plot.

qqnorm(x) − normal probability plot of data in x.

qqline(x) − adds a line to a normal probability plot passing through 1Q and 3Q

Probability distribution computations:

dbinom(x, n, p) − P(X = x) where X ∼ B(n, p)

pnorm(x, mean, sd) − P(X < x) where X ∼ N(mean, sd)

qnorm(p, mean, sd) − the value of x in p = P(X < x), where X ∼ N(mean, sd)

pt(x, df) − P(X < x) where X ∼ t(df)

qt(p, df) − the value of x in p = P(T < x), where T ∼ t(df)

pchisq(x, df) − P(X 2 < x) where X 2 ∼ χ2 (df)

Random sampling (without replacement):

sample(n) − a random arrangement of the first n positive integers.

sample(n, size) − a random sample of size values from among the first n positive integers.

2
Statistical inference:

t.test(x, mu) − one-sample t-test or confidence interval with data in x, with null hypothesized
value mu.

t.test(x1, x2) − two-sample t-test or confidence interval for difference in means with data in
x1 and x2

t.test(y ~ x, data=data.df) − alternative method for two-sample t-test; y is the quantitative


response and x is binary categorical variable in data frame data.df.

prop.test(x, n, p) − one-sample z-test or confidence interval for a Binomial probability, with


x successes in a sample size of n, and a hypothesized probability p.

prop.test(x, n) − two-sample z-test or confidence interval for difference in Binomial probabili-


ties, with x containing two counts of successes, and n containing two sample sizes.

mcnemar.test(x) − McNemar’s test for difference in Binomial probabilities with paired data,
with x containing 2 × 2 data frame.

aov(y ~ x, data=data.df) − analysis of variance of response y on categorical variable x con-


tained in data frame data.df.

lm(y~x1+x2+x3+..., data=data.df) − least-squares regression of y on x1, x2, etc., within data


frame data.df.

glm(y~x1+x2+x3+..., family=binomial, data=data.df) − logistic regression of y on x1, x2,


etc., within data frame data.df.

summary(model.fit) − summarize model.fit, the results of either analysis of variance, least-


squares regression, or logistic regression.

step(model.fit) − stepwise variable selection for least-squares or logistic regressions, with largest
model in model.fit.

predict(model.fit, newdata=newdata.df) − prediction of least-squares or logistic regression


model in model.fit using data in newdata.df.

fitted(model.fit) − fitted values from model.fit.

residuals(model.fit) − residuals from model.fit.

chisq.test(x, p) − chi-squared goodness-of-fit test, with vector of counts in x and vector of


probabilities in p.

chisq.test(x) − chi-squared test of independence, with counts in x as a data frame.

You might also like