組合投資研究-之-r軟件語言

統(tǒng)計軟件和R語言裝了R沒有？R免費R

資源公開(不是黑盒子,也不是

鬼)R可以在UNIX,Windows和Macintosh運行.R

有優(yōu)秀的內(nèi)在幫助系統(tǒng).R有優(yōu)秀的畫圖功能學(xué)生能夠輕松地轉(zhuǎn)到商業(yè)支持的

S-Plus程序(如果需要使用商業(yè)

)R語言有一個強大的,容易學(xué)習(xí)的語法,有許多內(nèi)在的統(tǒng)計函數(shù).通過用戶自編程序,R語言很容易延伸和擴大.它就是這樣成長的.R是計算機編程語言.類似于UNIX語言,C語言,Pascal,Gauss語言等.對于熟練的編程者,它將覺得該語言比其他語言更熟悉.而對計算機初學(xué)者,學(xué)習(xí)R語言使得學(xué)習(xí)下一步的其他編程不那么

.那些傻瓜

(SAS,SPSS等)語言的語法則完全不同.R的缺點沒有商業(yè)支持(但有網(wǎng)上支持)需要編程,不傻瓜.速度不如C++或FORTRANR的歷史S語言在1980年代后期在AT&T開發(fā).R項目由Auckland大學(xué)統(tǒng)計系的Robert

Gentleman和RossIhaka于1995年開始的.它很快得到廣泛用戶的歡迎.目發(fā)展團隊維持;它組成的工作努力前它是由R是一個由的國際團隊點擊CRAN得到一批鏡像R(

)點擊鏡像比如Berkeley選擇這個,安裝文件選擇這個,包選擇basePackages

（每個都有大量數(shù)據(jù)和可以讀寫修改的函數(shù)/程序）base

The

R

Base

Packageboot

Bootstrap

R

(S-Plus)

Functions

(Canty)class

Functions

for

Classificationcluster

Cluster ysis

Extended

Rousseeuw

et

al.concord

Concordance

and

reliabilitydatasets

The

R

Datasets

PackageexactRankTests

ExactDistributions

for

Rank

and

PermutationTestsforeign

Read

Data

Stored

by

Minitab,

S,

SAS,

SPSS,

Stata,

Systat,

dBase,

...graphics

The

R

Graphics

PackagegrDevices

The

R

Graphics

Devices

and

Support

for

Colours

and

Fontsgrid

The

Grid

Graphics

PackageKernSmooth

Functions

for

kernel

smoothing

for

Wand

&

Jones

(1995)lattice

Lattice

Graphics

Interfacetools

Tools

for

Package

Developmentutils

The

R

Utils

PackagePackages

（繼續(xù)）ysisMASSMain

Package

of

Venables

andRipley's

MASSmethodsFormal

Methods

and

ClassesmgcvGAMswith

GCV

smoothness

estimation

and

GAMMs

by

REML/PQLmulttestResampling-based

multiple

hypothesistestingnlmeLinear

and

nonlinearmixed

effects

modelsnnetFeed-forward

Neural

Networks

and

Multinomial

Log-Linear

ModelsnortestTests

for

NormalityoutliersTests

for

outliersplsPartial

Least

Squares

Regression

(PLSR)

and

Principal

Component

Regression

(PCR)pls.pcrPLS

and

PCR

functionsrpartRecursive

PartitioningSAGxStatistical ysisof

t

eChipsmaStatisticalMicroarray

ysisspatialFunctions

for

Kriging

and

Point

PatternsplinesRegression

Spline

Functions

and

ClassesstatsThe

R

Stats

Packagestats4Statistical

Functions

using

S4

ClassessurvivalSurvival ysis,

including

penalisedlikelihood.tcltkTcl/Tk

InterfacetoolsTools

for

Package

DevelopmentutilsTheR

Utils

PackagePackages

（網(wǎng)上）網(wǎng)上還有許多所有這些Packages都是在base

和stats

package上添加的Base,stats包含所有固有的應(yīng)用和數(shù)據(jù)而其他的packages包含各統(tǒng)計學(xué)家自己發(fā)展的方法和數(shù)據(jù)。希望你是下一個加盟這些packages的作者之一。賦值和運算z=

rnorm(1000000,4,0.1)median(z)賦值:“=”可以用“<-”代替x<-z->y->w簡單數(shù)

算有:+,-,*,/,

^,%*%,%%(mod)%/%(整數(shù)除法)等等常用的數(shù)學(xué)函數(shù)有:abs,sign,log,log2,log10,logb,expm1,

log1p(x),sqrt,exp,sin,cos,

tan,acos,asin,atan,cosh,sinh，tanh還有round, floor,

ceilinggamma

,

lgamma,

digamma

and

trigamma.sum,

prod,

cumsum,

cumprodmax,

min,

cummax,

cummin,

pmax,

pmin,rangemean,

length,

var,

duplicated,

uniqueunion,

intersect,

setdiff>,

>=,

<,

<=,&,

|,

!還有輸入輸出數(shù)據(jù):scan,read.table,

dump,save,load,

write,write.table,read.csvletters,

LETTERSlist,

matrix,

array,

cbind,

rbind,

mergesort,

order,

sort.list,

rev,

stack,

unstack

,reshape序列和向量z=seq(-1,10,length=100)z=seq(-1,10,

len=100)z=seq(10,-1,-1)z=10:-1x=rep(1:3,3)x=rep(3:5,1:3)>

x[1]

3

4

4

55

5x=rep(c(1,10),c(4,5))w=c(1,3,x,z);w[3]分布和產(chǎn)生隨機數(shù)正態(tài)分布:pnorm(1.2,2,1);dnorm(1.2,2,1);qnorm(.7,2,1);rnorm(10,0,1)#rnorm(10)t分布:pt(1.2,1);dt(1.2,2);qt(.7,1);rt(10,1)此外還有指數(shù)分布、F分布、“卡方”分布

Beta

分布、二項分布、Cauchy

分布、

Gamma分布、幾何分布、超幾何分布、對數(shù)正態(tài)分布、Logistic分布、負二項分布、

Poisson

分布、均勻分布、Weibull分布、

Willcoxon分布等變元可以是向量!向量運算x=rep(0,10);z=1:3;x+z[1]

1

2

3

1

2

3

1

2

3

1Warning

message:

longerobject

length

is

not

a

multiple

of

shorter

object

length

in:

x

+

zx*z[1]

0

0

0

0

0

0

0

0

0

0Warning

message:

longerobject

length

is

not

a

multiple

of

shorter

object

length

in:

x

*zrev(x)z=c("no

cat","has

","nine","tails")z[1]=="no

cat"[1]

TRUE向量名字和appendx=1:3;names(x)=LETTERS[1:3]xA

B

C12

3append(x,runif(3),after=2)A

B

C1.0000000

2.0000000

0.3107987

0.7505149

0.5752226

3.0000000向量賦值z=1:5z[7]=8;z[1]

1

2

3

4 5

NA

8z=NULLz[c(1,3,5)]=1:3;z[1] 1

NA 2

NA

3rnorm(10)[c(2,5)]z[-c(1,3)]#去掉第1、3元素.z[(length(z)-4):length(z)]#最后五個元素.向量的大小次序z=sample(1:100,10);z#比較sample(1:100,10,rep=T)[1]

75

68

28

42

17

21

96

34

69

47order(z)[1]

5

6

3

8 4

10

2

9

1

7z[order(z)][1]

17

21

28

34

42

47

68

69

75

96sort(z)[1]

17

21

28

34

42

47

68

69

75

96which(z==max(z))#給出下標Matrix[2,]

2

610

14

18[3,]

3

711

15

19[4,]

4

812

16

20x=matrix(runif(20),4,5)>

x[,1]

[,2]

[,3]

[,4]

[,5][1,]

0.7983678

0.04607601

0.04555323

0.8594483

0.73089500[2,]

0.6559851

0.79562222

0.02948270

0.1453364

0.79552838[3,]

0.6759171

0.56193147

0.48286653

0.2419931

0.56069988[4,]

0.1183701

0.80652627

0.49405167

0.6523137

0.08345406>

x=matrix(1:20,4,5);x[,1]

[,2]

[,3]

[,4][,5][1,]

1

5

9

13

17>

x=matrix(1:20,4,5,byrow=T);x[,1]

[,2]

[,3]

[,4]

[,5][1,]

1

2

3

4

56

7

8

9

10[2,][3,][4,]11

12

13

14

1516

17

18

19

20一些簡單函數(shù)max,min,length,mean,median,fivenum,

le,unique,sd,var,range,rep,diff,sort,order,sum,cumsum,prod,cumprod,rev,print,sample,seq,exp,pi矩陣的行和列(子集)nrow(x);

ncol(x);dim(x)#行列數(shù)目x=matrix(rnorm(24),4,6)x[c(2,1),]#第2和第1行x[,c(1,3)]#第1和第3列x[2,1]

#第[2,1]元素x[x[,1]>0,1]

#第1列大于0的元素sum(x[,1]>0)#第1列大于0的元素的個數(shù)sum(x[,1]<=0)#第1列不大于0的元素的個數(shù)x[,-c(1,3)]#沒有第1、3列的x.x[-2,-c(1,3)]

#沒有第2行、第1、3列的x.矩陣/向量的(子集)x[x[,1]>0&x[,3]<=1,1];

#第1中大于0并且相應(yīng)于第3列中小于或等于1的元素(“與”)x[x[,2]>0|x[,1]<.51,1]

#第1中小于.51或者相應(yīng)于第2列中大于0的元素(“或”)x[!x[,2]<.51,1]#第一列中相應(yīng)于第2列中不小于.51的元素(“非”)邏輯運算:>,<,==,<=,>=,!=;&,|,!x=rnorm(10)all(x>0);all(x!=0);any(x>0);(1:10)[x>0]x=sample(1:7,5,rep=T);unique(x)矩陣的轉(zhuǎn)置和逆矩陣x=matrix(runif(9),3,3);x[,1]

[,2]

[,3][1,]

0.6747652

0.9954731

0.7524502[2,]

0.3090199

0.2390141

0.2472961[3,]

0.5102675

0.9515505

0.6082803t(x)[,1]

[,2]

[,3][1,]

0.6747652

0.3090199

0.5102675[2,]

0.9954731

0.2390141

0.9515505[3,]

0.7524502

0.2472961

0.6082803solve(x)

# solve(a,b)可以解ax=b方程[,1]

[,2]

[,3][1,]

-12.313293

15.125819

9.082300[2,]

-8.459725

3.627898

8.989864[3,]

23.563034

-18.363808

-20.037986警告:計算機中的0是什么?x%*%solve(x)[,1]

[,2]

[,3][1,]

1.000000e+00

-9.454243e-17

-3.911801e-16[2,]

5.494737e-16

1.000000e+00

3.248270e-16[3,]

-3.018419e-16

1.804980e-15

1.000000e+00要用線性代數(shù)的知識來判斷諸多少非零特征根等問題.假定v是特征值組成的向量,不能用諸如

sum(v!=0)等方法來判斷非零特征根的數(shù)目!Matrix

&

Arrayx=array(runif(20),c(4,5));

x[,1]

[,2]

[,3]

[,4]

[,5][1,]

0.5474306

0.2362356

0.687007107

0.4036998

0.5255839[2,]

0.8234363

0.4922711

0.960554564

0.4704976

0.1327870[3,]

0.1861151

0.8461655

0.390523424

0.2202575

0.4057607[4,]

0.8117521

0.5375946

0.004505845

0.4821567

0.7644741is.matrix(x)[1]TRUEx[1,2]x[1,]x[,2]dim(x)#得到維數(shù)(4,5)Arrayx=array(runif(24),c(4,3,2))is.matrix(x)

#可由dim(x)得到維數(shù)(4,3,2)[1]

FALSEx,,

1[,1]

[,2]

[,3][1,]

0.3512615

0.7270611

0.009055522[2,]

0.1444965

0.2527673

0.697977027[3,]

0.6658176

0.6638542

0.773747542[4,]

0.4258436

0.4168940

0.634235148,

,

2[,1]

[,2]

[,3][1,]

0.3664152

0.9633497

0.5628006[2,]

0.3466645

0.5036830

0.1542986[3,]

0.4552553

0.1289775

0.8423017[4,]

0.1074899

0.3841463

0.7648297Array的子集>x=array(1:24,c(4,3,2))x[c(1,3),,],

,

1[,1]

[,2]

[,3][1,]159[2,]3711,,

2[,1]

[,2]

[,3][1,]

13

17

21[2,]

15

19

23矩陣乘法及行列運算[3,]

-1.750057-0.02764783

1.694761

3.4171705.1395786.861987[4,]

5.8624129.78064218

13.698872

17.61710321.53533325.453563x=matrix(1:30,5,6);y=matrix(rnorm(20),4,5)y%*%x[,1]

[,2]

[,3]

[,4]

[,5]

[,6][1,]

-3.231808

-8.13791204

-13.044017

-17.950121

-22.856225

-27.762330[2,]

-14.072030

-39.33640851

-64.600787

-89.865165

-115.129543

-140.393921apply(x,1,mean)[1]

13.5

14.5

15.5

16.5

17.5apply(x,2,sum)[1]

140apply(x,2,prod)[1]

120

30240 360360

1860480

6375600

17100720Array的維運算x=array(1:24,c(4,3,2))apply(x,1,mean)[1]

11

12

13

14apply(x,1:2,sum)[,1]

[,2]

[,3][1,]

14

22

30[2,]

16

24

32[3,]

18

26

34[4,]

20

28

36apply(x,c(1,3),prod)[,1][,2][1,] 454641[2,]120

5544[3,]2316555[4,]384

7680矩陣與向量之間的運算sweep(x,1,1:5,"*")[,1]

[,2]

[,3][,4]

[,5][,6][1,]16111621

26[2,]414243444

54[3,]924395469

84[4,]1636567696

116[5,]255075100125

150x*1:5sweep(x,2,1:6,"+")[,1]

[,2]

[,3]

[,4]

[,5]

[,6][1,]2814202632[2,]3915212733[3,]41016222834[4,]51117232935[5,]61218243036scale(x):Standardizes

the

elementsof

x,

i.e.subtracts

the

mean

and

divides

by

thestandard

deviation.

Results

in

a

vector

withzero

mean

and

unit

standard

deviation.Also

works

with

data

frames

(column-wiseand

only

with

numeric

data!).update.packages()complete.casesalgae

<-na.omit(algae).RDatasave.image()存的函數(shù)例子se

<-

function(x)

{v

<-

var(x)n

<-

length(x)return(sqrt(v/n))}函數(shù)例子basic.stats

<-

function(x,more=F)

{stats

<-

list()clean.x

<-

x[!is.na(x)]stats$n

<-

length(x)stats$nNAs

<-

stats$n-length(clean.x)stats$mean

<-

mean(clean.x)

stats$std

<-

sd(clean.x)stats$med

<-median(clean.x)if

(more){stats$skew<-

sum(((clean.x-stats$mean)/stats$std)^3)/length(clean.x)stats$kurt

<-

sum(((clean.x-stats$mean)/stats$std)^4)/length(clean.x)

-3}stats}basic.stats(c(45,2,4,46,43,65,NA,6,-213,-3,-45))Array和矩陣/向量/array之間的運算z=array(1:24,c(2,3,4))#注意排列次序z,

,

1[,1]

[,2][,3][1,]

1

3

5[2,]

2

4

6,

,

2[,1]

[,2][,3][1,]

7

9

11[2,]

8

10

12,

,

3[,1]

[,2][,3][1,]

13

15

17[2,]

14

16

18,

,

4[,1]

[,2][,3][1,]

19

21

23[2,]

20

22

24Array和矩陣/向量/array之間的運算sweep(z,1,1:2,"-"),

,

1[,1]

[,2]

[,3][1,]

0

2

4[2,]

0

2

4,

,

2[,1]

[,2]

[,3][1,]

6

8

10[2,]

6

8

10,

,

3[,1]

[,2]

[,3][1,]

12

14

16[2,]

12

14

16,

,

4[,1]

[,2]

[,3][1,]

18

20

22[2,]

18

20

22Array和矩陣/向量/array之間的運算sweep(z,c(1,2),matrix(1:6,2,3),"-"),

,

1[,1]

[,2]

[,3][1,]

0

0

0[2,]

0

0

0,

,

2[,1]

[,2]

[,3][1,]

6

6

6[2,]

6

6

6,

,

3[,1]

[,2]

[,3][1,]

12

12

12[2,]

12

12

12,

,

4[,1]

[,2]

[,3][1,]

18

18

18[2,]

18

18

18外積(產(chǎn)生矩陣或array)[1,]11[2,]22outer(1:2,rep(1,2))[,1]

[,2]outer(1:2,matrix(rep(1,6),3,2)),

,

1[,1]

[,2]

[,3][1,]111[2,]222,,

2[,1]

[,2]

[,3][1,]111[2,]222List(set

of

objects)list可以是任何對象的集合(包括lists)z=list(1:3,Tom=c(1:2,

a=list("R",letters[1:5]),w="hi!"))z[[1]];z[[2]];z$T;z$T$a2;z$T[[3]];z$T$wattributes(z)#屬性!$names""

"Tom"attributes(matrix(1:6,2,3))$dim[1]

23矩陣,array及其維名字x=matrix(1:12,nrow=3,dimnames=list(c("I","II","III"),paste("X",1:4,sep="")))X1

X2

X3

X4I

1

4 7

10II

2

5 8

11III

3

6 9

12y=array(1:12,c(3,2,2),dimnames=list(c("I","II","III"),paste("X",1:2sep=""),paste("Y",1:2,

sep=""))),

,

Y1X1

X21

42

53

6,

,

Y2X1

X2I 7

108

119

12data.framex=matrix(1:6,2,3)x=as.data.frame(x);xV1

V2

V31

1 3

52

2 4

6x$V2[1]

3

4x$V2[1]

3

4attributes(x)$names[1]

"V1"

"V2"

"V3"$s[1]

"1"

"2"$class[1]

"data.frame"data.frame20011352002246names(x)=c("TOYOTA","GM","HUNDA")s(x)=c("2001","2002")xTOYOTA

GM

HUNDAx$GM[1]

3

4data.frameattach(x)GM[1]

3

4detach(x)GMError:

Object

"GM"

not

found直接手工輸入和編輯數(shù)據(jù)直接敲入:x=c(1,2,7,8,…)或者x=scan()1

2

7

8….(以“Enter”兩次來結(jié)束)fix(x)(通過編輯修改數(shù)據(jù))Categorical

dataA

survey

asks

people

if

they

smoke

or

not.

Thedata

is

Yes,

No,

No,

Yes,

Yesx=c("Yes","No","No","Yes","Yes")table(x);xfactor(x)Barplot:Suppose,

a

group

of

25

people

are

surveyed

as

to

their

beer-drinking

preference.

The

categories

were

(1)

Domestic

can,(2)Domesticbottle,

(3)

Microbrew

and

(4)

import.

The

raw

data

is

3

4

12

32

3

1

1

1

1

4

3

1beer

=

scan()3

4

1 2

3

2

3

1

1

1

1

4

3

1barplot(beer)

#

this

isn't

correctbarplot(table(beer))

#

Yes,call

with

summarized

databarplot(table(beer)/length(beer))

#

divide

by

n

for

proportiontable(beer)/length(beer)CEO

salaries:

Suppose,

CEO

yearlycompensations

aresampled

and

the

following

are

found

(in

millions).(This

isbefore

being

indicted

for

cooking

the

books.)

12

.4

5

2

50

83

1

4

0.25sals

=

scan()

#

read

in

with

scan12

.4

5

2

50

8

3

1

4

0.25mean(sals)

;var(sals)

;

sd(sals)

;median(sals)fivenum(sals)

#

min,

lower

hinge,

Median,

upper

hinge,

maxsummary(sals)d (10,

17,

18,

25,

28,28);

summary(data);le(data,.25);

le(d

(.25,.75))sort(sals);

fivenum(sals);summary(sals)mean(sals,trim=1/10)

;mean(sals,trim=2/10)IQR(sals)Mad:median|Xi-median(X)|(1.4826)mad(sals)median(abs(sals

-

median(sals)))

#

withoutmedian(abs(sals

-

median(sals)))

*

1.4826Stem-and-leaf

Charts

Suppose

you

have

the

boxscore

of

a

basketball

game

and

the

following

pointsper

game

for

players

on

both

teams2

0

0

0

6

28

31

14

4

8

2

52

3scores

=

scan()2

32

0

0

0

6

28

31

14

4

8

2

5apropos("stem")#`apropos‘returns

a

charactervector

giving

the

names

of

all

objects

in

the

searchlist

matching

`what’.如>apropos(“stem”)[1]

“stem”

“system”

“system.file”“system.time”

參看find("stem")stem(scores);stem(scores,scale=2)The

salaries

could

beplaced

into

broad

categories

of

0-1million,

1-5

million

and

over

5

million.

To

do

this

usingRone

uses

the

cut()function

and

the

table()

function.Suppose

the

salaries

are

again12

.45

2

50

8

3

1

4

.25

And

we

want

to

break

that

dataintothe

intervals

[0;

1];

(1;

5];

(5;

50]sals

=

c(12,

.4,

5,

2,

50,

8,

3,

1,

4,

.25)

#

enter

datacats

=

cut(sals,breaks=c(0,1,5,max(sals)))

#

the

breakscats

#

view

the

valuestable(cats)

#

organizelevels(cats)

=

c("poor","rich","rolling

in

it")table(cats)Histograms:

Suppose

the

top

25

rankedmovies

madethefollowing

gross

receipts

for

a

week

429.6

28.2

19.6

13.7

13.0

7.8

3.4

2.0

1.9

1.0

0.7

0.4

0.4

0.3

0.30.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.1x=scan()29.6

28.2

19.6

13.7

13.0

7.8

3.4

2.0

1.9

1.0

0.7

0.4

0.4

0.3

0.30.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.1hist(x)

#

frequencieshist(x,probability=TRUE)

#

proportions(or

probabilities)rug(jitter(x))

#

add

tick

markshist(x,breaks=10)

#

10

breaks,

or

just

hist(x,10)hist(x,breaks=c(0,1,2,3,4,5,10,20,max(x)))

#

breaksFrequency

Polygons:x=c(.314,.289,.282,.279,.275,.267,.266,.265,.256,.250,.249,.211,.161)tmp

=

hist(x)

#

store

the

resultslines(c(min(tmp$breaks),tmp$mids,max(tmp$breaks)),c(0,tmp$counts,0),type="l")data(faithful)attach(faithful)

#

make

eruptions

visiblehist(eruptions,15,prob=T)

#

proportions,

not

frequencieslines(density(eruptions))

#

lines

makes

a

curve,

default

bandwidthHandling

bivariate

categorical

data:Smokes

amount

of

StudyingY

less

than

5

hoursN

5

-

10

hoursN

5

-

10hoursY

more

than

10

hoursN

more

than

10

hoursY

less

than

5

hoursY

5

-

10hoursY

less

than

5

hoursN

more

than

5

hoursY

5

-

10

hourslibrary(MASS)quineattach(quine)table(Age)table(Sex,

Age);

tab=xtabs(~

Sex

+

Age,quine);

unclass(tab)tapply(Days,

Age,

mean)tapply(Days,

list(Sex,

Age),

mean)apply,

sapply,

tapply,

lapplysmokes

=

c("Y","N","N","Y","N","Y","Y","Y","N","Y")amount

=

c(1,2,2,3,3,1,2,1,3,2)table(smokes,amount)tmp=table(smokes,amount)

#

store

the

tableoptions(digits=3)

#

only

print

3

decimal

placesprop.table(tmp,1)

#

the

rows

sum

to

1

nowprop.table(tmp,2)

#

the

columns

sum

to

1

nowReally

`sweep(x,

margin,

margin.table(x,

margin),

"/")‘prop.table(tmp)#amount

#

all

the

numbers

sum

to

1options(digits=7)

#

restore

the

number

of

digitsPlotting

tabular

databarplot(table(smokes,amount))barplot(table(amount,smokes))smokes=factor(smokes)

#

for

namesbarplot(table(smokes,amount),

beside=TRUE,legend.text=T)barplot(table(amount,smokes),main="table(amount,smokes)",

beside=TRUE,

legend.text=c("lessthan

5","5-10","more

than

10"))categorical

vs.

numerical:A

simple

example

mightbe

in

adrug

test,

where

you

have

data

(in

suitable

units)

for

anexperimental

group

and

for

a

control

group.experimental:

5

5

5

13

7

11

11

9

8

9control:

11

8

4

5

9

5

10

5

4

10x

=

c(5,

5,

5,

13,

7,

11,

11,

9,

8,9)y

=

c(11,

8,

4,

5,

9,

5,

10,

5,

4,

10)boxplot(x,y);amount

=

scan()5

5

5

13

711

11

9

8

9

11

8

4

5

9

5

10

5

4

10category

=

scan()1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2boxplot(amount

~

category)

#

note

the

tilde

~從文本文件輸入ASCII碼數(shù)據(jù)x=scan("f:\\book\\1.txt")這是按照文本一行一行讀入的數(shù)據(jù)如果原先是4

5矩陣形式,則加用x=matrix(x,4,5,byrow=T)或直接用x=

matrix(scan("f:\\book\\1.txt"),4,5,b=T)如果原先是4

5有名字的data.frame形式,則用x=read.table("f:\\book\\ww.txt",header=T)xGM

VW

HUNDA6

7

8控制語句x=NULL;for(iin

1:5)x=cbind(x,i^2)i=1;x=NULL;while(i<=5){x=cbind(x,i^2);i=i+1}x=rnorm(1);if

(x>0)

y=x

else

y=-x+10i=1;x=rnorm(1);repeat{x=x+rnorm(1);if(x>3)break;i=i+1};print(c(i,x))怎么調(diào)出Packages來使用?Packages:libraries敲library(),就知道有什么libraries了,缺省library是base.比如要進入mass,就敲>library(MASS)每個library都有許多數(shù)據(jù)在每個library敲data(),就知道有什么數(shù)據(jù)了比如敲data(Titanic),就調(diào)出數(shù)據(jù)Titanic來了.注意:R語言對大小寫敏感.怎么從網(wǎng)上Packages來使用?從R上網(wǎng)直接到CRAN主頁zip文件到你的計算機用本機的zip來安裝程序包…byattach(warpbreaks)by(warpbreaks[,

1:2],

tension,

summary)by(warpbreaks[,

1],

list(wool

=

wool,

tension

=

tension),

summary)##

now

suppose

we

wantto

extract

the

coefficients

by

grouptmp

<-by(warpbreaks,

tension,

function(x)

lm(breaks

~

wool,

data

=

x))sapply(tmp,

coef)%in%和match##

The

intersection

of

two

sets

:intersect

<-

function(x,

y)

y[match(x,

y,

nomatch

=

0)]intersect(1:10,7:20)1:10%in%c(1,3,5,9)sstr

<-

c("c","ab","B","bba","c","@","bla","a","Ba","%")sstr[sstr

%in%

c(letters,LETTERS)]"%w/o%"

<-

function(x,y)

x[!x

%in%

y]

#-- x

without

y(1:10)

%w/o%c(3,7,12)intersect

<-

function(x,

y)

y[match(x,y,

nomatch

=

0)]intersect(1:10,7:20)attach(warpbreaks)warpbreaks[tension%in%

c("L","H"),]warpbreaks[warpbreaks$tension%in%

c("L","H"),]warpbreaks[warpbreaks[,3]%in%

c("L","H"),]unique(tension)ftable：把array/矩陣（沒有頻率的）數(shù)據(jù)變成列聯(lián)表（找出計數(shù)）##

Start

with

a

contingencytable.ftable(Titanic,

row.vars

=

1:3)ftable(Titanic,

row.vars

=

1:2,

col.vars

=

“Survived”)ftable(Titanic,

row.vars=

2:1,

col.vars

=

“Survived”)##

Start

with

a

data

frame.x

<-

ftable(mtcars[c(“cyl”,

“vs”,

“am”,

“gear”)])x#為array，其維的次序為(“cyl”,“vs”,“am”,“gear”)ftable(x,row.vars=c(2,4))#從x（array）確定表的行變量##Start

with

expressions,use

table()'s"dnn"to

change

labelsftable(mtcars$cyl,

mtcars$vs,

mtcars$am,

mtcars$gear,

row.vars

=

c(2,

4),dnn

=

c("Cylinders",

"V/S",

"Transmission",

"Gears"))ftable(vs~carb,mtcars)#vs是列，carb是行#或ftable(mtcars$vs~mtcars$carb)ftable(carb~vs,mtcars)

#vs是行，carb是列

ftable(mtcars[,c(8,11)])#和上面ftable(carb~vs,mtcars)等價

ftable(breaks~wool+tension,warpbreaks)as.data.frame(UCBAdmissions)把array（三維列聯(lián)表）變成方陣(DF

<-as.data.frame(UCBAdmissions))Admit

Gender

Dept

FreqAdmitted

Male A

512Rejected

Male A

313………xtabs(Freq

~

Gender

+

Admit,

DF)把有頻率/計數(shù)的方陣變成列聯(lián)表xtabs(Freq

~

Admit+

Gender

+

Dept,

DF)把方陣變成原來的列聯(lián)表(a=xtabs(Freq~

Admit+Gender,data=DF))#如無頻數(shù)(權(quán))，左邊為空Gender

Admit Male

FemaleAdmitted

1198

557Rejected

1493

1278library(MASS);biplot(corresp(a,

nf=2))#應(yīng)用之一的互換上一頁數(shù)據(jù)中變量的屬性和定量is.factor(DF[,2])[1]

TRUEis.factor(DF[,3])[1]

TRUEis.factor(DF[,4])[1]

FALSEDF[,4]=as.factor(DF[,4])is.factor(DF[,4])[1]

TRUEDF[,4]=as.numeric(DF[,4])is.factor(DF[,4])[1]

FALSE，則可能出錯在用啞元記錄屬性變量觀測時，如不改變mtcars[1:4,]mpg

cyl

disp

hp

drat wt

qsec

vs

amgear

carbMazdaRX4 21.0

6

160

110

3.90

2.620

16.46

0

1

4

4Mazda

RX4

Wag

21.0

6

160

1103.90

2.875

17.02

0

1

4

4Datsun

710 22.8

4

108

93

3.85

2.320

18.61

1

1

4

1Hornet

4

Drive

21.4

6

258

1103.08

3.215

19.44

1

0

3

1lm(mpg~gear+carb,mtcars)

#把定性變量當成定量變量Call:lm(formula

=

mpg~

gear

+

carb,

data

=mtcars)Coefficients:(Intercept)7.276gear

carb5.576 -2.754mtcars[,10]=as.factor(mtcars[,10])#改變mtcars[,11]=as.factor(mtcars[,11])#改變(lm(mpg~gear+carb,mtcars))Call:lm(formula

=

mpg

~

gear

+

carb,

data

=mtcars)Coefficients:(Intercept)gear4gear5carb2carb3carb420.9327.7208.349-3.289-4.632-9.064carb6-9.581carb8-14.281向量比較：allx=1:12;y=1:12;all(y==x)[1]

TRUEcat

和printif(all(x

<

0))

cat("all

x

values

arenegative\n")all

x

values

are

negativecat("a

logical

or

(positive)

numeric

controlling

how

the

outputis++broken

into

successive

lines. If

'FALSE'

(default),

onlynewlines

created

explicitly

by\n")a

logical

or

(positive)

numeric

controlling

how

the

output

isbroken

into

successivelines. If

'FALSE'

(default),

onlynewlines

created

explicitly

byif(all(x

<0))

print("all

x

values

arenegative\n")[1]

"all

x

values

arenegative\n"print("a

logical

or

(positive)

numeric

controlling

how

the

output

is++broken

into

successive

lines. If

'FALSE'

(default),

onlynewlines

created

explicitly

by")[1]

"a

logical

or

(positive)

numeric

controlling

how

the

output

is\nbroken

into

successivelines. If

'FALSE'

(default),

only\nnewlines

created

explicitly

by"解釋后面語句的意思n=1200;x=runif(n*10)x=matrix(x,n,10)x1=(x[,1:5]<.4)*1p=x1*.8+(1-x1)*.2x2=1*(x[,6:10]<p)z=cbind(x1,x2)ss=function(z){nu=NULL;pattern=NULLz0

=as.matrix(z);n=nrow(z0);id=1;repeat{if

(id%%100==0)print(id)z1=sweep(z0,2,z0[1,],"==")pattern=rbind(pattern,z0[1,])z2=apply(z1,1,prod)z0=z0[z2!=1,];nu=c(nu,sum(z2))if

(sum(nu)>=n)breakif

(is.matrix(z0)==F){nu=c(nu,1);pattern=rbind(pattern,z0);id=id+1break}id=id+1}list(pattern=pattern,number=nu,id=id)}R-語言畫圖x=0:18plot(x,x,pch=x,col=x)points(x,18-x,pch=x)matplot(x,cbind(x,18-x))畫圖spring=data.frame(compression=c(41,39,43,53,42,48,47,46),distance=c(120,114,132,157,122,144,137,141))attach(spring)(Hooke’s

law:

f=.5ks)plot(distance

~

compression)plot(compression,

distance)畫圖par(mfrow=c(2,2))#準備畫2

2的4個圖plot(compression,

distance,main=

"Hooke'sLaw")#只有標題的圖plot(compression,

distance,main=

"Hooke'sLaw",xlab=

"x",ylab="y")#標題+x,y標記identify(compression,distance)#標出點號碼plot(compression,

distance,main="Hooke'sLaw")#只有標題的圖text(46,120,

"f=1/2*k*s")#在指定位寫入文字plot(compression,

distance,main="Hooke'sLaw")#只有標題的圖text(locator(2),

"I

am

here!")

#在點擊的兩個位置寫入文字畫圖library(mass);data(Animals);attach(Animals)plot(body,

brain)plot(sqrt(body),plot((body)^0.1,sqrt(brain))(brain)^0.1)plot(log(body),log(brain))0200004000060000800000100030005000bodybrain0501001502002503000204060sqrt(body)sqrt(brain)1.01.52.02.53.01.01.52.0(body)^0.1(brain)^0.1051002468log(body)log(brain)畫圖(原始數(shù)據(jù))par(mfrow=c(1,1),

pch=1)plot(body,

brain,

xlim=c(0,

6000))text(x=body,y=brain,labels=s(Animals),adj=0)#adj=0

implies

left

adjusted

textplot(log(body),

log(brain))text(x=log(body),y=log(brain),labels=s(Animals),adj=0)#0100020003000400050006000010002000300040005000bodybrainAsian

elephantHumanoknekyeymesatverrHGoirasfefe

ChDGimopnriaklClenaozyeweR畫圖(幾個點)plot(body[c(1,3,8)],brain[c(1,3,8)],xlim=c(0,200))text(x=body[c(1,3,8)],y=brain[c(1,3,8)],labels=s(Animals[,c(1,3,8)]),

adj=0)0501001502000100200300400body[c(1,

3,

8)]brain[c(1,

3,

8)]Mount