PackageSoft/R-033U.tex (2018/March) R:

Size: px

Start display at page:

Download "PackageSoft/R-033U.tex (2018/March) R:"

ちえこいさやま
4 years ago
Views:

................................................................................ R: 2018 3 29................................................................................ R AI R https://cran.

1 R: R AI R R R Mase-Rstatman.pdf (i) (ii) 1 R 1.1 RGui(64bit) R Console R - RStudio RGui(64bit) (working deirectry) setwd("") getwd() _ 1/9

2 , Enter Shift + Enter column xdat < - c(1,2,3,4,5,6) > sum(xdat) > [1] 21 [1] <- > xdat2 = c(10:20); sum(xdat2) > [1] 165 version = 165 ( ) > xfdat = read.csv("c:\\rfiles\\rdat123.txt", + header = FALSE,sep=",") c: xfdat > > xfdat > xout = c(10,11,12,13,14,15,16,17,18,19,20,21,22) > write(xout, "c:\\rfiles\\rdat124.txt", sep=" ", ncolumns= 5) > write(c("mean(xout)=", mean(xout)), "c:\\rfiles\\rdat124.txt", + append=true) mean(xout)= (append) Excel read.tabel write.table data.frame > outdat = data.frame( + = c( 101, 102, 103, 104, 105), + = c(173, 172, 168, 170, 174)) > write.table(outdat,"\\rfiles\\smp.csv",sep=",",row.names=false) 2/9

read.table header = TRUE 101, 101, outdat[,1] outdat[,2] 173, 172, 1 101 173 2 102 172 3 103 168 4 104 170 5 105 174 1.2 plot(xdat,ydat, ) type = l ( ) :, type = p :, type = h :.

3 read.table header = TRUE 101, 101, outdat[,1] outdat[,2] 173, 172, plot(xdat,ydat, ) type = l ( ) :, type = p :, type = h :. xlabel = x, ylabel = y, main =, curve(expr,from,to,option), add = TRUE R Graphics Device - - > curve(sin(x),-2*pi,2*pi,xlabel= x, ylabel= y,main= y=sin(x),cos(x) ) > # sin(x), x > curve(cos(x),-2*pi,2*pi,add=true) # cos(x) y=sin(x),cos(x) y x f(x) f(x) = 1 2π e x2 /2, < x < R f(x, y) 3/9

4 0 < ρ < 1 f(x, y) = 1 2π 1 ρ 2 exp { (x 2 2ρxy + y 2 )/(2(1 ρ 2 )) }, < x, y < ρ = 0 x, y > x <- seq(-3,3,length=50) # x > y <- x # y > rho <- 0.9 # 2 > gauss3d <- function(x,y) { # 2 + 1/(2*pi*sqrt(1-rho^2))*exp(-(x^2-2*rho*x*y+y^2) / (2*(1-rho^2)))} > z <- outer(x,y,gauss3d) # z > z[is.na(z)] <- 1 # 1 > persp(x, y, z, theta = 30, phi = 30, expand = 0.5, col = "lightblue") persp(x, y, z, col =,, phi =, expand =, border=na) theta = 1.3 mean(, na.rm=true); # na.rm=true (NA) (remove) medain( ); range( ); max) - (min) xdat > xdat=scan() Enter Enter (histogram) > xdat =rnom(100) # 100 xdat > hist(xdat, 6, xlab= rnorm(100), ylab= Frequency, main = Histogram of rnorm(100) ) > # xdat 6 ( ) > hist( ) ( 0, 1) 4/9

5 =NORM.INV(RAND(), ) R Adata = scan(".csv") (boxplot) 1000 > A1000 = matrix(runif(1e3), nrow=100, ncol=10) > # unif 1e3 = 1*10^3 = 1000 > dfa = as.data.frame(a1000) > # data.frame dfa > boxplot(dfa) # ( ) > adat=rnorm(20) # 20 + bdat=rnorm(20) # 20 + cdat=rnorm(100) # ddat=rnorm(100) # edat=rnorm(1000) # boxplot(adat, bdat, cdat, ddat, edat) # xdat:, ydat:, name.arg =, ylim = (bar chart) : barplot(ydat,name.arg=xdat, ylim=c(0,200)) (pie chart) : pie(datname, radious=size, col=, main= title name ) # names (xdata) =c( name1, name2, name3, name4, name5 ) (stem and leaf): 5/9

6 > xdat = c( ) > stem(xdat, scale=1) 3 The decimal point is 1 digits(s) to the right of the hist() (plot.histogram) right=true a 1 < x a 2, a 2 < x a 3, a 3 < x a 4, a 1 a 1 (i) (Sturges) (ii) (Scott) (iii) FD (Freedman-Diaconis) breaks = Sturges (i) n! n C k = k!(n k)! k = 0, 1, 2,, n 2n h 1 n h n = h 1C i = 2 h 1 h 1 = log 2 n 1 + log 2 n (ii) n ˆσ(x) h = 7 ˆσ(x) 2 3 n (iii) Q 1, Q 3 Interquantile IQR(x) = Q 3 Q 1 h = 2 IQR(x) 3 n 7 ˆσ(x) 2IQR(x) 2 i=0 2.2 Binom(n,p) ; (random number of binomial distribution) rbinom(12, size=5, p=0.8) () Uniform(0,1) ; (random number of uniform distribution) runif(120) 120 [0,1] N(0,1); (random number of standard normal distribution) rnorm(120) 0, 1 ( 1) 120 x ϕ(x; µ, σ 2 ) = 1 e (x µ)2 /2σ 2 (probability density) d 2πσ dnorm(x, mean = µ, sd = σ) q qnorm(x) mean = µ = 0, sd= σ = 1 (distribution function) Φ(x) = P (X x) = x ϕ(t; 0, 1)dt (exponential); [r] rexp(n=5, rate=λ) : P (X = t) = λe λt, t > 0 (geometric); rgeom(n=5, prob=0.2) : P (X = k) = (1 p) k p, k = 0, 1, 2, 6/9

7 p = 0.2 k p, 1 p (Poisson); rpois(n=5, lambda=3) : P (X = k) = λk k! e λ, k = 0, 1, 2, λ = 3 5 (i) t (Student ) (ii) (chi-square) (iii) F (Snedecor ) (distribution) d (i) >dnorm(x, mean =0, sd =1) ; mean sd x ( ) (n+1)/2 Γ((n + 1)/2) (ii) >dt(x, df) ; df f(x; n) = 1 + x2, < x < 2nΓ(n/2) n (iii) 1 >dchisq(x, df) ; df = n, f(x; n) = 2 n/2 Γ(n/2) xn/2 1 e x/2, x > 0 (iv) >df(x, df1, df2) ; df1 = n 1, df2= n 2, f(x; n 1, n 2 ) = Γ((n 1 + n 2 )/2) Γ(n 1 /2)Γ(n 2 /2) ( n1 n 2 ) n1/2 ( x n1/ n ) (n1+n 2)/2 1 x, x > 0 n 2 3 R nrow(xdata) xdat (row), ncol(xdata) xdat (column) rownames(xdata) xdat colnames(xdata) xdat, > mata + matb # a_{i,j} + b_{i,j} > mata - matb # a_{i,j} - b_{i,j} > mata / matb # a_{i,j} / b_{i,j} > mata %*% matb # \sum_j \{a_{i,j} * b_{j,k}\}, > solve(mata) # Inverse(matA) # xdata > xdata <-c(12, 14, 25, 53, 6, 91, 25, 89, 77, 34) # > source("c:.r") 7/9

8 mean(xdata) var(xdata) sd(xdata) median(xdata) sum(xdata) max(xdata) min(xdata) rev(xdata) order(xdata) sort(xdata) rev(sort(xdata)) xdata 3.2 <- function( ){ } sekiwa <- function(xdata, ydata){sum(xdat * ydata)} # sum(xdata[i] * ydata[i]) dat2[5] dat2[,3] (c ) \verb+ cor(dat2[,2], dat2[,3]) + b (correlatin) \verb+ sum(dat2[,1]) + a (summation) \includegraphics[width=8cm]{r3.eps} %*% 8/9

9 org/wiki/r 9/9

講義のーと : データ解析のための統計モデリング. 第２回

講義のーと : データ解析のための統計モデリング. 第２回 Title 講義のーと : データ解析のための統計モデリング Author(s) 久保, 拓弥 Issue Date 2008 Doc URL http://hdl.handle.net/2115/49477 Type learningobject Note この講義資料は, 著者のホームページ http://hosho.ees.hokudai.ac.jp/~kub ードできます Note(URL)http://hosho.ees.hokudai.ac.jp/~kubo/ce/EesLecture20