Download Table of Useful R Commands Cheat Sheet and more Cheat Sheet Probability and Statistics in PDF only on Docsity!
Table of Useful
R
commands
Command
Purpose
help()
Obtain documentation for a given
R
command
example()
View some examples on the use of a command
c()
,^ scan()
Enter data manually to a vector in
R
seq()
Make arithmetic progression vector
rep()
Make vector of repeated values
data()
Load (often into a data.frame) built-in dataset
View()
View dataset in a spreadsheet-type format
str()
Display internal structure of an R object
read.csv()
,^ read.table()
Load into a data.frame an existing data file
library()
,^ require()
Make available an
R
add-on package
dim()
See dimensions (# of rows/cols) of data.frame
length()
Give length of a vector
ls()
Lists memory contents
rm()
Removes an item from memory
names()
Lists names of variables in a data.frame
hist()
Command for producing a histogram
histogram()
Lattice command for producing a histogram
stem()
Make a stem plot
table()
List all values of a variable with frequencies
xtabs()
Cross-tabulation tables using formulas
mosaicplot()
Make a mosaic plot
cut()
Groups values of a variable into larger bins
mean()
,^ median()
Identify “center” of distribution
by()
apply function to a column split by factors
summary()
Display 5-number summary and mean
var()
,^ sd()
Find variance, sd of values in vector
sum()
Add up all values in a vector
quantile()
Find the position of a quantile in a dataset
barplot()
Produces a bar graph
barchart()
Lattice
command for producing bar graphs
boxplot()
Produces a boxplot
bwplot()
Lattice
command for producing boxplots
Command
Purpose
plot()
Produces a scatterplot
xyplot()
Lattice
command for producing a scatterplot
lm()
Determine the least-squares regression line
anova()
Analysis of variance (can use on results of
lm()
predict()
Obtain predicted values from linear model
nls()
estimate parameters of a nonlinear model
residuals()
gives
(observed - predicted)
for a model fit to data
sample()
take a sample from a vector of data
replicate()
repeat some process a set number of times
cumsum()
produce running total of values for input vector
ecdf()
builds empirical cumulative distribution function
dbinom()
, etc.
tools for binomial distributions
dpois()
, etc.
tools for Poisson distributions
pnorm()
, etc.
tools for normal distributions
qt()
, etc.
tools for student
t^
distributions
pchisq()
, etc.
tools for chi-square distributions
binom.test()
hypothesis test and confidence interval for 1 proportion
prop.test()
inference for 1 proportion using normal approx.
chisq.test()
carries out a chi-square test
fisher.test()
Fisher test for contingency table
t.test()
student t test for inference on population mean
qqnorm()
,^ qqline()
tools for checking normality
addmargins()
adds marginal sums to an existing table
prop.table()
compute proportions from a contingency table
par()
query and edit graphical settings
power.t.test()
power calculations for 1- and 2-sample
t
anova()
compute analysis of variance table for fitted model
Examples of usage
help()
help(mean)
example()
require(lattice) example(histogram)
c(), rep() seq()
x = c(8, 6, 7, 5, 3, 0, 9) x [1] 8 6 7 5 3 0 9
names = c("Owen", "Luke", "Anakin", "Leia", "Jacen", "Jaina") names [1] "Owen" "Luke" "Anakin" "Leia" "Jacen" "Jaina" heartDeck = c(rep(1, 13), rep(0, 39)) heartDeck [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 [49] 0 0 0 0
y = seq(7, 41, 1.5) y [1] 7.0 8.5 10.0 11.5 13.0 14.5 16.0 17.5 19.0 20.5 22.0 23.5 25.0 26.5 28.0 29.5 31.0 32.5 34. [20] 35.5 37.0 38.5 40.
data(), dim(), names(), View(), str()
data(iris) names(iris) [1] "Sepal.Length" "Sepal.Width" "Petal.Length" "Petal.Width" "Species" dim(iris) [1] 150 5 str(iris) 'data.frame': 150 obs. of 5 variables: $ Sepal.Length: num 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4. ... $ Sepal.Width : num 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3. ... $ Petal.Length: num 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1. ... $ Petal.Width : num 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0. ... $ Species : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ... View(iris)
library(), require()
library(abd) require(lattice)
histogram()
require(lattice) data(iris) histogram(iris$Sepal.Length, breaks=seq(4,8,.25)) histogram(~ Sepal.Length, data=iris, main="Iris Sepals", xlab="Length") histogram(~ Sepal.Length | Species, data=iris, col="red") histogram(~ Sepal.Length | Species, data=iris, n=15, layout=c(1,3))
read.csv()
As.in.H2O = read.csv("http://www.calvin.edu/~scofield/data/comma/arsenicInWater.csv")
read.table()
senate = read.table("http://www.calvin.edu/~scofield/data/tab/rc/senate99.dat", sep="\t", header=T)
mean(), median(), summary(), var(), sd(), quantile(),
counties=read.csv("http://www.calvin.edu/~stob/data/counties.csv") names(counties) [1] "County" "State" "Population" "HousingUnits" "TotalArea" [6] "WaterArea" "LandArea" "DensityPop" "DensityHousing" x = counties$LandArea mean(x, na.rm = T) [1] 1126. median(x, na.rm = T)
[1] 616.
summary(x) Min. 1st Qu. Median Mean 3rd Qu. Max. 1.99 431.70 616.50 1126.00 923.20 145900. sd(x, na.rm = T) [1] 3622. var(x, na.rm = T) [1] 13122165
quantile(x, probs=seq(0, 1, .2), na.rm=T) 0% 20% 40% 60% 80% 100% 1.99 403.29 554.36 717.94 1043.82 145899.
sum()
firstTwentyIntegers = 1: sum(firstTwentyIntegers) [1] 210 die = 1: manyRolls = sample(die, 100, replace=T) sixFreq = sum(manyRolls == 6) sixFreq / 100 [1] 0.
stem()
monarchs = read.csv("http://www.calvin.edu/~scofield/data/comma/monarchReigns.csv") stem(monarchs$years) The decimal point is 1 digit(s) to the right of the |
0 | 0123566799 1 | 0023333579 2 | 012224455 3 | 355589 4 | 4 5 | 069 6 | 3
table(), table(), mosaicplot(), cut()
pol = read.csv("http://www.calvin.edu/~stob/data/csbv.csv") table(pol$sex) Female Male 133 88 table(pol$sex, pol$Political04) Conservative Far Right Liberal Middle-of-the-road Female 67 0 14 48 Male 47 7 6 28 xtabs(~sex, data=pol) sex
Female Male 133 88
xtabs(~Political04 + Political07, data=pol) Political Political04 Conservative Far Left Far Right Liberal Middle-of-the-road Conservative 58 0 2 13 39 Far Right 4 0 3 0 0 Liberal 0 1 1 14 4 Middle-of-the-road 20 0 0 22 32 mosaicplot(~Political04 + sex, data=pol)
barplot()
pol = read.csv("http://www.calvin.edu/~stob/data/csbv.csv") barplot(table(pol$Political04), main="Political Leanings, Calvin Freshman 2004") barplot(table(pol$Political04), horiz=T) barplot(table(pol$Political04),col=c("red","green","blue","orange")) barplot(table(pol$Political04),col=c("red","green","blue","orange"), names=c("Conservative","Far Right","Liberal","Centrist"))
Conservative Liberal Centrist
0
40
80
barplot(xtabs(~sex + Political04, data=pol), legend=c("Female","Male"), beside=T)
Conservative Far Right Liberal Middle−of−the−road
Female Male
0
10
20
30
40
50
60
boxplot()
data(iris) boxplot(iris$Sepal.Length) boxplot(iris$Sepal.Length, col="yellow") boxplot(Sepal.Length ~ Species, data=iris) boxplot(Sepal.Length ~ Species, data=iris, col="yellow", ylab="Sepal length",main="Iris Sepal Length by Species")
l
setosa versicolor virginica
Iris Sepal Length by Species
Sepal length
plot()
data(faithful) plot(waiting~eruptions,data=faithful) plot(waiting~eruptions,data=faithful,cex=.5) plot(waiting~eruptions,data=faithful,pch=6) plot(waiting~eruptions,data=faithful,pch=19) plot(waiting~eruptions,data=faithful,cex=.5,pch=19,col="blue") plot(waiting~eruptions, data=faithful, cex=.5, pch=19, col="blue", main="Old Faithful Eruptions", ylab="Wait time between eruptions", xlab="Duration of eruption")
l
l
l l
l
l
l l
l
l
l
l l
l
l
l
l
l
l
l
ll
l l
l
l
l
l (^) l l ll l
l l
l l
l
l
l l
l
l
l
l
l
l l
l
l
l
l
l
l
l
l l l
l l
l
l
l
l
l
l
l l l
l
l
l
l l l
l
l
l l
l l
l l l
l
l l l
l
l
l
l
l
l
l
l
l l
l
l
l
l
l
l l
l
l
l
l l l
l
l l
l
l
l
l
l
l
l
l
l
l
l l
l
l
l
l
l
l
l
l
l
l
l
l
l
l l
l
l l l
l
l
l
l
l
l l l
l l (^) l
l
l
l
l
l
l
l
l l
l l
l
l
l
l
l
l l
l l l
l
l l
l
l
l l
l
l
l
l
l
l
l
l
l
l l l
l l
l
l
l
l
l
l
l
l
l
l
l
l l
l
l
l l
l
l
l
l
l
l
l
l
l ll l l l
l l
l
l
l
l
ll
l l
l
l
l
l
l
l l
l
l
l
l
l
l l l
l l l
l
l
l (^) l
l
l
l
l
l
l
l
l
l
l
l
Old Faithful Eruptions
Duration of eruption
Time between eruptions
dbinom(), pbinom(), qbinom(), rbinom(), binom.test(), prop.test()
dbinom(0, 5, .5) # probability of 0 heads in 5 flips
[1] 0.
dbinom(0:5, 5, .5) # full probability dist. for 5 flips [1] 0.03125 0.15625 0.31250 0.31250 0.15625 0. sum(dbinom(0:2, 5, .5)) # probability of 2 or fewer heads in 5 flips [1] 0. pbinom(2, 5, .5) # same as last line [1] 0. flip5 = replicate(10000, sum(sample(c("H","T"), 5, rep=T)=="H")) table(flip5) / 10000 # distribution (simulated) of count of heads in 5 flips flip 0 1 2 3 4 5 0.0310 0.1545 0.3117 0.3166 0.1566 0. table(rbinom(10000, 5, .5)) / 10000 # shorter version of previous 2 lines 0 1 2 3 4 5 0.0304 0.1587 0.3087 0.3075 0.1634 0.
qbinom(seq(0,1,.2), 50, .2) # approx. 0/.2/.4/.6/.8/1-quantiles in Binom(50,.2) distribution [1] 0 8 9 11 12 50 binom.test(29, 200, .21) # inference on sample with 29 successes in 200 trials Exact binomial test
data: 29 and 200 number of successes = 29, number of trials = 200, p-value = 0. alternative hypothesis: true probability of success is not equal to 0. 95 percent confidence interval: 0.09930862 0. sample estimates: probability of success
prop.test(29, 200, .21) # inference on same sample, using normal approx. to binomial
1-sample proportions test with continuity correction
data: 29 out of 200, null probability 0. X-squared = 4.7092, df = 1, p-value = 0. alternative hypothesis: true p is not equal to 0. 95 percent confidence interval: 0.1007793 0. sample estimates: p
pchisq(), qchisq(), chisq.test()
1 - pchisq(3.1309, 5) # gives P-value associated with X-squared stat 3.1309 when df=
[1] 0.
pchisq(3.1309, df=5, lower.tail=F) # same as above [1] 0. qchisq(c(.001,.005,.01,.025,.05,.95,.975,.99,.995,.999), 2) # gives critical values like Table A [1] 0.002001001 0.010025084 0.020100672 0.050635616 0.102586589 5.991464547 7. [8] 9.210340372 10.596634733 13. qchisq(c(.999,.995,.99,.975,.95,.05,.025,.01,.005,.001), 2, lower.tail=F) # same as above [1] 0.002001001 0.010025084 0.020100672 0.050635616 0.102586589 5.991464547 7. [8] 9.210340372 10.596634733 13. observedCounts = c(35, 27, 33, 40, 47, 51) claimedProbabilities = c(.13, .13, .14, .16, .24, .20) chisq.test(observedCounts, p=claimedProbabilities) # goodness-of-fit test, assumes df = n- Chi-squared test for given probabilities
data: observedCounts X-squared = 3.1309, df = 5, p-value = 0.
addmargins()
blood = read.csv("http://www.calvin.edu/~scofield/data/comma/blood.csv") t = table(blood$Rh, blood$type) addmargins(t) # to add both row/column totals A AB B O Sum Neg 6 1 2 7 16 Pos 34 3 9 38 84 Sum 40 4 11 45 100 addmargins(t, 1) # to add only column totals A AB B O Neg 6 1 2 7 Pos 34 3 9 38 Sum 40 4 11 45 addmargins(t, 2) # to add only row totals
A AB B O Sum Neg 6 1 2 7 16 Pos 34 3 9 38 84
par()
par(mfrow = c(1,2)) # set figure so next two plots appear side-by-side poisSamp = rpois(50, 3) # Draw sample of size 50 from Pois(3) maxX = max(poisSamp) # will help in setting horizontal plotting region hist(poisSamp, freq=F, breaks=-.5:(maxX+.5), col="green", xlab="Sampled values") plot(0:maxX, dpois(0:maxX, 3), type="h", ylim=c(0,.25), col="blue", main="Probabilities for Pois(3)")
Histogram of poisSamp
Sampled values
Density
Probabilities for Pois(3)
0:maxX
dpois(0:maxX, 3)
fisher.test()
blood = read.csv("http://www.calvin.edu/~scofield/data/comma/blood.csv") tblood = xtabs(~Rh + type, data=blood) tblood # contingency table for blood type and Rh factor type Rh A AB B O Neg 6 1 2 7 Pos 34 3 9 38 chisq.test(tblood) Pearson's Chi-squared test
data: tblood X-squared = 0.3164, df = 3, p-value = 0.
fisher.test(tblood) Fisher's Exact Test for Count Data
data: tblood p-value = 0. alternative hypothesis: two.sided
dpois(), ppois()
dpois(2:7, 4.2) # probabilities of 2, 3, 4, 5, 6 or 7 successes in Pois(4.211)
[1] 0.13226099 0.18516538 0.19442365 0.16331587 0.11432111 0.
ppois(1, 4.2) # probability of 1 or fewer successes in Pois(4.2); same as sum(dpois(0:1, 4.2)) [1] 0. 1 - ppois(7, 4.2) # probability of 8 or more successes in Pois(4.2) [1] 0.
pnorm() qnorm(), rnorm(), dnorm()
pnorm(17, 19, 3) # gives Prob[X < 17], when X ~ Norm(19, 3) [1] 0. qnorm(c(.95, .975, .995)) # obtain z* critical values for 90, 95, 99% CIs
[1] 1.644854 1.959964 2.
nSamp = rnorm(10000, 7, 1.5) # draw random sample from Norm(7, 1.5) hist(nSamp, freq=F, col="green", main="Sampled values and population density curve") xs = seq(2, 12, .05) lines(xs, dnorm(xs, 7, 1.5), lwd=2, col="blue")
Sampled values and population density curve
nSamp
Density
t.test()
data(sleep) t.test(extra ~ group, data=sleep) # 2-sample t with group id column Welch Two Sample t-test
data: extra by group t = -1.8608, df = 17.776, p-value = 0. alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: -3.3654832 0. sample estimates: mean in group 1 mean in group 2 0.75 2.
sleepGrp1 = sleep$extra[sleep$group==1] sleepGrp2 = sleep$extra[sleep$group==2] t.test(sleepGrp1, sleepGrp2, conf.level=.99) # 2-sample t, data in separate vectors Welch Two Sample t-test
data: sleepGrp1 and sleepGrp t = -1.8608, df = 17.776, p-value = 0. alternative hypothesis: true difference in means is not equal to 0 99 percent confidence interval: -4.027633 0. sample estimates: mean of x mean of y 0.75 2.
qqnorm(), qqline()
qqnorm(precip, ylab = "Precipitation [in/yr] for 70 US cities", pch=19, cex=.6) qqline(precip) # Is this line helpful? Is it the one you would eyeball?
l l
l
l
l l
l l
ll l
l
l l
l l
llll l
l
l
lll llll
l ll
l
l
l
l
l
l
l l
l ll
l
l ll l ll^ l l
l
l
l l l
l
l
l
l
ll l
l
l l l
l
Normal Q−Q Plot
Precipitation [in/yr] for 70 US cities Theoretical Quantiles
power.t.test()
power.t.test(n=20, delta=.1, sd=.4, sig.level=.05) # tells how much power at these settings
Two-sample t test power calculation
n = 20 delta = 0. sd = 0. sig.level = 0. power = 0. alternative = two.sided
NOTE: n is number in each group
power.t.test(delta=.1, sd=.4, sig.level=.05, power=.8) # tells sample size needed for desired power Two-sample t test power calculation
n = 252. delta = 0. sd = 0. sig.level = 0. power = 0. alternative = two.sided
NOTE: n is number in each group
anova()
require(lattice) require(abd) data(JetLagKnees) xyplot(shift ~ treatment, JetLagKnees, type=c('p','a'), col="navy", pch=19, cex=.5) anova( lm( shift ~ treatment, JetLagKnees ) ) Analysis of Variance Table
Response: shift Df Sum Sq Mean Sq F value Pr(>F) treatment 2 7.2245 3.6122 7.2894 0.004472 ** Residuals 19 9.4153 0.
Signif. codes: 0 ***' 0.001**' 0.01 *' 0.05.' 0.1 ` ' 1
treatment
shift −
−
0
control eyes knee
lll llll l
ll ll ll l
ll lll l l
