1 R Windows R 1.1 R The R project web R web Download [CRAN] CRAN Mirrors Japan Download and Install R [Windows 9

Size: px

Start display at page:

Download "1 R Windows R 1.1 R The R project web R web Download [CRAN] CRAN Mirrors Japan Download and Install R [Windows 9"

みちしげうなだ
4 years ago
Views:

1 1 R Windows R 1.1 R The R project web R web Download [CRAN] CRAN Mirrors Japan Download and Install R [Windows 95 and later ] [base] R for Windows R [R win32.exe] 1.2 R Windows R [ ] [OK] R web 1

2 1.3 R web 1. RjpWiki 2. R-Tips 3. R 4. R R R mjin/r/index.html R URL1 web URL2 R 1 URL3 web 1 URL4 R PDF 2 R 2.1 R R Microsoft Excel R R > R 2

3 > # [1] -52 > 3 * 90 / 45 # [1] 6 > sqrt(2) # sqrt() [1] # 2 sqrt() 5 R > 12 ^ 2 # [1] 144 > sin(1) #sin" " [1] > cos(5) #cos" " [1] > tan(3) #tan" " [1] > log10(12) # ( 10 ) [1] > exp(1) # [1] R 3

4 +, -, *, / 2.2 R > a <- 244 # a > a # [1] 244 > b < # b > b [1] > c <- "Katasho" # c > c [1] "Katasho" 1 1 R 4

5 > a <- c(1, 2, 3, 4, 5) # > a [1] > b <- c("a", "b", "c", "d", "e") # > b [1] "a" "b" "c" "d" "e" > c <- c(1, "b", 3, "d", 5) # > c [1] "1" "b" "3" "d" "5" 1 c() 1 c a a [1, 2, 3, 4, 5] a [1, 2, 3, 4, 5] a

6 > dat <- matrix(c( + 1, 2, 3, + 4, 5, 6, + 7, 8, 9), nrow=3, ncol=3, byrow=t) > dat [,1] [,2] [,3] [1,] [2,] [3,] > dat2 <- matrix(c( + 1, 2, 3, + 4, 5, 6, + 7, 8, 9), ncol=3) > dat2 [,1] [,2] [,3] [1,] [2,] [3,] matrix() 2 matrix() nrow=3 3 ncol= ncol byrow byrow=t dat2 1 6

7 > dat4 <- cbind(x1, x2, x3) # cbind() > dat4 x1 x2 x3 [1,] [2,] [3,] > dat5 <- rbind(x1, x2, x3) # rbind() > dat5 [,1] [,2] [,3] x x x

8 > my.data <- data.frame(x1 = x1, X2 = x2, Y = x3) # > my.data X1 X2 Y > attach(my.data) #my.data > X1 [1] > X2 [1] > Y [1] > detach(my.data) # > X1 "X1" attach() detach() detach() 2.3 R 8

9 > dat <- rnorm(100, mean=50, sd=10) > mean(dat) # [1] > median(dat) # [1] > var(dat) # [1] > sd(dat) # [1] > dat2 <- rnorm(100, mean=100, sd=20) > var(dat, dat2) # [1] > cor(dat, dat2) # [1] t t t R R t t.test() paired paired=false t t var.equal=false 9

10 > group1 <- c(56, 49, 59, 42, 58, 61, 53) # > group2 <- c(73, 86, 54, 60, 65, 70, 53) # > t.test(group1, group2, var.equal=false) # t Welch Two Sample t-test data: group1 and group2 t = , df = 9.518, p-value = alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: sample estimates: mean of x mean of y > t.test(group1, group2, paired=true) # t Paired t-test data: group1 and group2 t = , df = 6, p-value = alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: sample estimates: mean of the differences summary() aov()

11 10 1 > #1 > class.one <- c(3.9, 3.8, 3.6, 4.4, 4.7, 3.5, 3.7, 4.6, 3.9, 3.6) > #2 > class.two <- c(5.8, 4.4, 4.5, 6.4, 4.9, 7.2, 4.8, 5.7, 5.6, 3.4) > #3 > class.thr <- c(6.0, 5.8, 6.1, 5.6, 6.2, 6.0, 6.1, 5.7, 6.4, 5.9) > group <- as.factor(rep(c(1, 2, 3), c(10, 10, 10))) # > # > my.data <- data.frame( + Y = c(class.one, class.two, class.thr), X = group) > result <- summary(aov(y ~ X, my.data)) # result > result # Df Sum Sq Mean Sq F value Pr(>F) X e-06 *** Residuals Signif. codes: 0 *** ** 0.01 * t R web web 11

12 4 R Excel > x <- rnorm(100) > y <- rnorm(100) > plot(x, y) # > plot(x, y, xlim=c(-3,3), ylim=c(-3,3)) #x y > plot(x, y, xlab=" ", ylab=" ") #x y > plot(x, y, col="red", pch=18) # (col) (pch) R plot() plot() plot() > x <- 1:6 > y <- c(1, 5, 6, 3, 9, 10) > plot(x, y, type="b") # > plot(x, y, type="b", lty=2) # plot() type type= l type= s type= n plot() barplot() 12

13 # 1 > x <- c(12, 34, 72, 56, 20) > names(x) <- c("a1", "A2", "A3", "A4", "A5") > barplot(x, col="blue") > abline(h=0) # 2 > y <- c(17, 34, 69, 32, 12) > mat <- rbind(x, y) > mat A1 A2 A3 A4 A5 x y > barplot(mat, beside=true, col=c("blue", "red")) > legend(2, 60, + paste(c(" "," ")), + fill=c("blue", "red")) # k 3 > g1 <- rnorm(10, 10, 1) > g2 <- rnorm(10, 20, 1) > g3 <- rnorm(10, 5, 1) > #x > plot(x, c(g1,g2,g3), xlab="group", ylab="score", xaxt="n") > axis(1, at=1:3, label=c("g1", "g2", "g3")) #x > M <- c(mean(g1), mean(g2), mean(g3)) # > x.jiku <- 1:3 # x > points(x.jiku, M, pch=3, col="red", type="b") # 13

14 5 5.1 (1) (2) (3) 3 mean() mean () ( ) { } 5.2 pow 1 pow(2) 2 2 = 4 # pow() pow <- function(x){ } res <- x ^ 2 res # > pow(2)#2 [1] 4 > pow(12) #12 [1]

15 # pow2 pow2 <- function(x, y){ res <- x ^ y res } # > pow2(2, 4) #2 4 [1] 16 > pow2(2, -4) #2-4 [1] R Windows R [ ][ ] R R sd() sd() } sd2 <- function(x){ var2 <- function(x){ } # x if(is.vector(x)) n <- length(x) # x if(is.matrix(x)) n <- nrow(x) var2 <- var(x) * (n-1) / n # sd2 <- sqrt(var2(x)) # var2 sd2 15

16 1 R R 5.3 R Excel R plot() #1 one.func <- function(x){ } 2*x + 1 # > plot(one.func) > thr.func <- function(x) 3*x^3-0.5*x^2 + sqrt(x) + (3/4) > plot(thr.func) plot() points() legend() lwd 16

17 > plot(thr.func, xlim=c(-2,2), ylim=c(-5,5)) #thr.func plot() > abline(h=0) # x > abline(v=0) # y > x.dat <- seq(-2, 2, 0.1) #one.func x > points(x.dat, 2*x.dat+1, type="l", col="pink") #points() > legend(-1.5, 4, + c("one.func", "thr.func"), lwd=2, + col=c("black", "pink")) # 17

「統計数学３」

「統計数学３」関数の使い方 1 関数と引数関数の構造関数名 ( 引数 1, 引数 2, 引数 3, ) 例 : マハラノビス距離を求める関数 mahalanobis(data,m,v) 引数名を指定して記述する場合 mahalanobis(x=data, center=m, cov=v) 2 関数についてのヘルプ基本的な関数のヘルプの呼び出し? 関数名例 :?mean 例 :?mahalanobis 指定できる引数を確認する関数