Microsoft PowerPoint - A02-hmwu_R-Object.pptx

Transcription

1 A02 R 程式語言的基礎 : 物件 吳漢銘國立臺北大學統計學系

2 本章大綱 & 學習目標 2/60 R 程式變數 (Variables) 命名法及語法 Rounding of Numbers 向量 (Vectors) 向量運算 規律的序列 邏輯向量 字元向量 遺失值 因子 (Factors) 陣列 (Arrays) 矩陣 (Matrices) 表列 (List) 資料框 (Data Frame) 時間日期 物件的模式 (Mode) 類別 (Class) 屬性 (Attributes) 查詢物件之模式 類別 屬性 結構及可取用的元素 存取物件內之元素

3 R 變數命名法 3/60 Case sensitive A and a are different All alphanumeric symbols are allowed (A-Z, a-z, 0-9)., _. Name must start with. or a letter. 錯誤命名 3x 3_x 3-x 3.x.3variable 正確命名 x_3 x3 x.3 taiwan.taipei.x3.variable Google's R Style Guide Function name: Make function names verbs. GOOD : CalculateAvgClicks BAD : calculate_avg_clicks BAD : calculateavgclicks Variable name: GOOD : avg.clicks OK: avgclicks BAD : avg_clicks File names: be meaningful. GOOD: predict_ad_revenue.r BAD: foo.r

4 R 語法 4/60 Assignments x <- 5 x = 5 x (x <- 5) [1] 5 assign("x", 2) 2 - x a <- b <- c <- 6 x <- expressions x <- 3+5 ; or new line x <- 5 ; y <- 7 or x <- 5 Y <- 7 Assignment: 建議使用 <-, 而不是 = Google's R Style Guide Comment # how are you? + : If a comment is not complete at the end of a line, R will give a different prompt. x <- + 5 (y <- 1:5) 提示符號 (prompt) [1] sum(y) [1] 15 options(prompt = 'hmwu ', continue = "+ ") hmwu sum(y^2) [1] 55 hmwu options(prompt = ' ', continue = "+ ")

5 物件 (Objects) 5/60 Variables, arrays of numbers, character strings, functions, x <- 3+5 y <- 7 objects() [1] "x" "y" ls() [1] "x" "y" rm(x, y) rm(list = ls()) objects() character(0) 儲存 R 物件所佔用的記憶體估計 : object.size(x) print(object.size(x), units = "Mb") n < p <- 200 mydata <- matrix(rnorm(n*p), ncol = p, nrow=n) print(object.size(mydata), units = "Mb") 15.3 Mb os <- function(x){ + print(object.size(x), units = "Mb") + } os(mydata) 15.3 Mb

6 物件 (Objects) 6/60 x <- 2 # x is a vector of length 1 x <- vector() # x is a vector of 0 length x <- matrix() # x is a matrix of 1 column, 1 row x <- 'Hello Dolly' # x is a vector containing 1 string x <- c('hello', 'Dolly') # x is a vector with 2 strings x <- function(){} # x is a function that does nothing The vectors are atomic objects all of their elements must be of the same mode. Most other types of objects in R are more complex than vectors. They may consist of collections of vectors, matrices, data frames and functions. When an object is created (for example with the assignment <-), R must allocate memory for the object. The amount of memory allocated depends on the mode of the object.

7 Rounding of Numbers 7/60 ceiling: the smallest integers not less than the corresponding elements of x. floor: the largest integers not greater than the corresponding elements of x. x n n+1 floor(x) ceiling(x) trunc: the integers formed by truncating the values in x toward 0. round: rounds the values in its first argument to the specified number of decimal places (default 0). (x <- c(pi, 1/3, -1/3, -pi)) [1] ceiling(x) [1] floor(x) [1] trunc(x) [1] round(x, 2) [1] round(x, 5) [1] # checking if a number is an integer (x <- c(1/3, 2, -4, 1.5)) [1] ceiling(x) == floor(x) [1] FALSE TRUE TRUE FALSE

8 設定數值顯示位數 (1) 8/60 getoption("digits") # digits: controls the number (1~22, default 7) of digits to print when printing numeric values. [1] 7 x1 <- c(-10, , 0, , 10) x2 <- c(-10, , 0, , 10, pi) x3 <- c(-12, , 0, , 12) x4 <- c( e+09, e-09, 10, pi, , ) x1 # same as print(x1) [1] -1e+01-1e-05 0e+00 1e-05 1e+01 x2 [1] x3 [1] x4 [1] e e e e e-07 cat(x1, "\n") -10-1e e cat(x2, "\n") -10-1e e cat(x3, "\n") cat(x4, "\n") e e-07

9 設定數值顯示位數 (2) 9/60 op <- options(); str(op) # nicer printing List of 72 $ add.smooth : logi TRUE... options(digits = 3) # try options(digits = 1) or options(digits = 5) x1 # same as print(x1) [1] -1e+01-1e-05 0e+00 1e-05 1e+01 x2 [1] x3 [1] x4 [1] 1.81e e e e e-07 cat(x1, "\n") -10-1e e cat(x2, "\n") -10-1e e cat(x3, "\n") cat(x4, "\n") 1.81e e e-07 options(op) # reset (all) initial options options("digits") # or getoption("digits") $digits [1] 7 See also: getoption( scipen ) # (penalty) print numeric values in fixed (scipen = positive integer) notation or exponential (negative integer) notation. Fixed notation will be preferred unless it is more than scipen digits wider. options(scipen = 100, digits = 4) options(scipen = 999) # do not show scientific notation

10 向量 (Vector) 10/60 向量 (vector): 同樣屬性 ( 數字或文字 ) 的資料的集合 c(): 串連多個字串 (combine values into a vector or list) x1 <- c(10, 5, 3, 6, 2.7) x1 [1] assign("x2", c(10, 5, 3, 6, 2.7)) x2 [1] c(10, 5, 3, 6, 2.7) - x3 x3 [1] length(x1) [1] 5 c(1,7:9) [1] c(1:5, 10.5, "next") [1] "1" "2" "3" "4" "5" "10.5" "next" x1[4] [1] 6 x1[2:4] [1] x1[c(4, 2, 1)] [1] x1[-3] [1] x1[x1<5] [1] x1[10] [1] NA x1[2] <- 32; x1 [1] x1[c(1, 3, 5)] <- c(1,2,3) x1 [1]

11 向量 (Vectors) 11/60 (y1 <- 1/x1) [1] length(y1) [1] 5 v1 <- x1 + y1+1; v1; length(v1) [1] [1] 5 (y2 <- c(x1, 0, x1)); length(y2) [1] [1] 11 length(x1) [1] 5 (v2 <- x1 + y2 + 1); length(v2) [1] Warning message: In x1 + y2 : longer object length is not a multiple of shorter object length [1] 11 x1 repeated 2.2 times, y2 repeated one time, 1 repeated 11 times

12 課堂練習 1 12/60 注意 : 打完一行程式, 就立刻執行, 查看結果

13 向量運算 (Vector Arithmetic) 13/60 一些簡單的數學計算 : +, -, *, /, ^ log(x), logb(x, b) pi, exp(x), sin(pi/2), cos(pi), tan(pi/4), abs(x), sqrt(x), length(x) prod(x) choose(n, k) factorial(x) 一些簡單的統計運算 : max(x), min(x) pmax(x), pmin(x) range(x) c(min(x), max(x)) mean(x) sum(x)/length(x) var(x), cov(x) sum((x-mean(x))^2)/(length(x)-1) sqrt(var(x)) median(x) summary(x) cor(x, y)?mean 其它函式應用查看 mean 的用法 sort(x)# 排序, 由小到大 mean(x, trim = 0, na.rm = FALSE,...) rank(x)# 排序等級 order(x)# 排序後, 各個元素的原始所在位置

14 課堂練習 2 14/60 x <- c(1.58, -0.29, 0.59, -0.38, 0.72) max(x) [1] 1.58 min(x) [1] range(x) [1] c(min(x), max(x)) [1] mean(x) [1] sum(x)/length(x) [1] var(x) [1] sum( (x-mean(x))^2)/(length(x)-1) [1] sqrt(var(x)) [1] median(x) [1] 0.59 summary(x) Min. 1st Qu. Median Mean 3rd Qu. Max sort(x) [1] rank(x) [1] order(x) [1]

15 規律的序列 (Regular Sequences) 15/60 x <- c(1,2,3,4,5,6,7,8,9,10) x <- 1:10 y <- 10:2 2*1:10 #The colon operator has high priority with an expression [1] n <- 10 1:n-1 [1] :(n-1) [1] width <- 1 seq(from = 2, to = 5, by = width) [1] :5 [1] s1 <- seq(-5, 5, by = 0.2) # s2 <- seq(length = 51, from = -5, by = 0.2) #

16 規律的序列 (Regular Sequences) 16/60 rep(x, times=5) [1] [19] [37] rep(x, each=5) [1] [19] [37] rep(1:4, each=5) [1] rep(letters[1:4], 3) [1] "A" "B" "C" "D" "A" "B" "C" "D" "A" "B" "C" "D" rep(letters[1:4], length.out=3) [1] "A" "B" "C"

17 課堂練習 3: 造出規律的序列 17/ Fibonacci number ( 可能需要用到 for): rev, sequence Ex: 將 [0, 2] 分成 20 等份的子區間 取左端點 取右端點 取子區間之中點 ([a, b] 之 partition, 用於 Riemann Sum)

18 邏輯向量 (Logical Vectors) 18/60 TRUE, FALSE T, F Logical operators <, <=,, =, ==,!= c1,c2: logical expression c1&c2: intersection and c1 c2: union or x <- c(12, 4, 7, 20, 13) x < 15 [1] TRUE TRUE TRUE FALSE TRUE x <= 15 [1] TRUE TRUE TRUE FALSE TRUE x 13 [1] FALSE FALSE FALSE TRUE FALSE x = 10 [1] TRUE FALSE FALSE TRUE TRUE x == 12 [1] TRUE FALSE FALSE FALSE FALSE x!= 20 [1] TRUE TRUE TRUE FALSE TRUE (x = 3) [1] 3 (x 10) [1] TRUE FALSE FALSE TRUE TRUE x!= 20 [1] TRUE TRUE TRUE FALSE TRUE (x 10) & (x!= 20) [1] TRUE FALSE FALSE FALSE TRUE (x 10) (x!= 20) [1] TRUE TRUE TRUE TRUE TRUE (x = 10) [1] TRUE FALSE FALSE TRUE TRUE 1*(x = 10) [1] (x = 15) [1] FALSE FALSE FALSE TRUE FALSE 2*(x = 15) [1]

19 遺失值 (Missing Values) 19/60 NA: not available, missing values z <- c(1:3, NA) z [1] NA ind <- is.na(z) ind [1] FALSE FALSE FALSE TRUE A vector of the same length as x all of whose values are NA x == NA [1] NA NA NA NA NA NaN: not a number, missing values 0/0 [1] NaN Inf - Inf [1] NaN See also: na.fail(x), na.pass(x), na.omit(x), na.exclude(x) is.na(xx)is TRUE both for NA and NaN values is.nan(xx)is only TRUE for NaNs

20 練習 : NA, NaN and Inf 20/60 x <- c(na, 0 / 0, Inf - Inf, Inf, 5) # Inf is a number. x [1] NA NaN NaN Inf 5 y <- data.frame(x, is.na(x), is.nan(x), x == Inf, x == 5) y x is.na.x. is.nan.x. x...inf x NA TRUE FALSE NA NA 2 NaN TRUE TRUE NA NA 3 NaN TRUE TRUE NA NA 4 Inf FALSE FALSE TRUE FALSE 5 5 FALSE FALSE FALSE TRUE colnames(y) <- c("x", "is.na(x)", "is.nan(x)", "x == Inf", "x == 5") y x is.na(x) is.nan(x) x == Inf x == 5 1 NA TRUE FALSE NA NA 2 NaN TRUE TRUE NA NA 3 NaN TRUE TRUE NA NA 4 Inf FALSE FALSE TRUE FALSE 5 5 FALSE FALSE FALSE TRUE

21 字元向量 (Character Vectors) 21/60 Character strings are entered using either double ( ) quotes or single ( ) quotes Character strings are printed using double quotes. (x <- "x-values") [1] "x-values" (y <- "New iteration results") [1] "New iteration results" (answer1 <- c("a1", "a2", "b1", "b3")) [1] "a1" "a2" "b1" "b3" (answer2 <- c('a1', 'a2', 'b1', 'b3')) [1] "a1" "a2" "b1" "b3" (answer3 <- c('a', "a2", 3)) [1] "a" "a2" "3" (answer4 <- c('my name is "Hank"')) [1] "My name is \"Hank\"" (answer5 <- c("my name is 'Hank'")) [1] "My name is 'Hank'" paste("a", 1:6, sep = "") [1] "A1" "A2" "A3" "A4" "A5" "A6" paste("today is", date()) [1] "Today is Wed Sep 24 13:26: " labs <- paste(c("x", "Y"), 1:10, sep = "") labs [1] "X1" "Y2" "X3" "Y4" "X5" "Y6" "X7" "Y8" "X9" "Y10"

22 跳脫字元 (Escape Character) 22/60 Escape Character: \n: new line. \t: tab. \b: backspace. cat("how are you?", "\n", "I'm fine.", "\n") How are you? I'm fine. cat("how are you?", "\t", "I'm fine.", "\n") How are you? I'm fine. cat("how are you?", "\b\b\b", "I'm fine.") How are yo I'm fine. NOTE: "\" is entered and printed as "\\" setwd("c:\\temp\\mydata")

23 索引向量 : Index Vector [] 23/60 A logical vector A vector of positive integral quantities x <- c(7, 2, 4, 9, NA, 4) x[2] [1] 2 x[5] [1] NA x[0] numeric(0) x[10] [1] NA y <- x[!is.na(x)] y [1] (x+1)[(!is.na(x))&(x0)] - z z [1] rep(c(1,2,2,1), times=3) [1] c("x","y")[rep(c(1,2,2,1), times=3)] [1] "x" "y" "y" "x" "x" "y" "y" "x" "x" "y" "y" "x"

24 索引向量 : Index Vector [] 24/60 A vector of negative integral quantities x <- c(7, 2, 4, 9, NA, 4) x[-2] [1] NA 4 x[-(1:3)] [1] 9 NA 4 A vector of character strings fruit <- c(5, 10, 1, 20) fruit [1] names(fruit) <- c("orange", "banana", "apple", "peach") fruit orange banana apple peach lunch <- fruit[c("apple", "orange")] lunch apple orange 1 5 x <- c(7, 2, 4, 9, NA, 4) x[is.na(x)] <- 0 x [1] y <- c(-7, 2, 4, 9, 0, -4) abs(y) [1] y[y<0] <- - y[y<0] y [1]

25 課堂練習 4 25/60 x <- c(a = 5, B = 3, third = 10) x A B third x[1] A 5 x["a"] A 5 x[c("third", "B")] third B 10 3 x[c(3, 1)] third A 10 5 names(x) [1] "A" "B" "third" names(x) <- c("aa", "BB", "CC") x AA BB CC # compare two "char" strings "A" < "B" [1] TRUE "Hank" <= "Tom" [1] TRUE "201" < "1" [1] FALSE c("201", "001") < "1" [1] FALSE TRUE c("1", "A", "a") < "a" [1] TRUE FALSE FALSE # compare T/F TRUE < FALSE [1] FALSE T F [1] TRUE # compare T/F, char, numeric 1 < "a" [1] TRUE "a" < FALSE [1] TRUE 1 T [1] FALSE 1 < T [1] FALSE 1 == T [1] TRUE 0 < F [1] FALSE 0 F [1] FALSE 0 == F [1] TRUE

26 因子 (Factors) 26/60 The levels of factors are stored in alphabetical order. scores <- c(60, 49, 90, 54, 54, 48, 61, 61, 51, 49, 49) gender <- c("f", "f", "m", "f", "m", "m", "m", "m", "f", "f", "m") levels(gender) NULL gender.f <- factor(gender) gender.f [1] f f m f m m m m f f m Levels: f m levels(gender.f) [1] "f" "m" table(gender.f) gender.f f m 5 6 levels(gender.f) <- c(" 女 ", " 男 ") gender.f [1] 女女男女男男男男女女男 Levels: 女男 (scores.mean <- tapply(scores, gender.f, mean)) 女男 grade <- as.factor(c("b", "F", "A", "C", "A", "C", "B", "A", "F", "D")) levels(grade) [1] "A" "B" "C" "D" "F" grade2 <- ordered(grade, levels = rev(levels(grade))) grade2 [1] B F A C A C B A F D Levels: F < D < C < B < A grade2[which(grade2 = "B")] [1] B A A B A Levels: F < D < C < B < A

27 因子 (Factors) 27/60 MyLetter <- c("c", "D", "A", "K", "A", "I", "J", "I", "K", "H", "A", "K", "K", "B", "E", "H", "G", "L", "H", "H", "I", "K", "B", "D") MyLetter.factor <- factor(myletter) MyLetter.factor [1] C D A K A I J I K H A K K B E H G L H H I K B D Levels: A B C D E G H I J K L table(myletter.factor) MyLetter.factor A B C D E G H I J K L MyLetter.ordered <- factor(myletter, levels = c("a", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L"), ordered = TRUE) MyLetter.ordered[1] < MyLetter.ordered[2] [1] TRUE table(myletter.ordered) MyLetter.ordered A B C D E F G H I J K L

28 陣列 (Arrays) 28/60 An array is a multiply subscripted collection of data entries. z <- 1:30 z [1] [29] dim(z) <- c(3,5,2) z,, 1 [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] ,, 2 [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] z[1,3,2] [1] 22 z[1,1,] [1] 1 16 z[1,,2] [1] z[1,1:2,1] [1] 1 4 z[-1,,],, 1 [,1] [,2] [,3] [,4] [,5] [1,] [2,] ,, 2 [,1] [,2] [,3] [,4] [,5] [1,] [2,]

29 陣列 (Arrays) 29/60 x <- array(1:20, dim=c(4,5)) x [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] [4,] i <- array(c(1:3, 3:1), dim=c(3,2)) i [,1] [,2] [1,] 1 3 [2,] 2 2 [3,] 3 1 x[i] <- 0 x [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] [4,] x <- c(1,2,3,4,5) x [1] z <- array(x, dim=c(3,4)) z [,1] [,2] [,3] [,4] [1,] [2,] [3,] t(z) #transpose [,1] [,2] [,3] [1,] [2,] [3,] [4,] See also: aperm {base}: Array Transposition

30 課堂練習 5: interval data 30/60 temperature January.a January.b February.a February.b AnQing BaoDing BeiJing BoKeTu ChangChun temparray <- array(0, dim=c(5,2,2)) temparray[,,1] <- as.matrix(temperature[,c(1,3)]) temparray[,,2] <- as.matrix(temperature[,c(2,4)]) temparray,, 1 [,1] [,2] [1,] [2,] [3,] [4,] [5,] ,, 2 [,1] [,2] [1,] [2,] [3,] [4,] [5,] colnames(temparray) <- c("january", "February") rownames(temparray) <- rownames(temperature) dimnames(temparray)[[3]] <- c("min", "max") dimnames(temparray) [[1]] [1] "AnQing" "BaoDing" "BeiJing" "BoKeTu" "ChangChun" [[2]] [1] "January" "February" [[3]] [1] "min" "max" temparray,, min January February AnQing BaoDing BeiJing BoKeTu ChangChun ,, max January February AnQing BaoDing BeiJing BoKeTu ChangChun

31 matrix(): 矩陣 31/60 A matrix is an array with two subscripts. x <- 1:20 A <- matrix(x, ncol=4) A [,1] [,2] [,3] [,4] [1,] [2,] [3,] [4,] [5,] A.1 <- matrix(x, ncol=4, byrow=true) A.1 [,1] [,2] [,3] [,4] [1,] [2,] [3,] [4,] [5,] nrow(a) [1] 5 ncol(a) [1] 4 dim(a) [1] 5 4 diag(a) [1] B <- matrix(x+2, ncol=4) A * B #element by element product [,1] [,2] [,3] [,4] [1,] [2,] [3,] [4,] [5,] A %*% t(b) #matrix product [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] [4,] [5,] x <- 4 diag(x) #identity matrix

32 矩陣 (Matrices) 32/60 apply(mat, 1, mean) # row means [1] apply(mat, 2, mean) # column means [1] apply(mat, 1, var) # row variances [1] apply(mat, 2, var) # column variances [1] mean(mat) [1] 10.5 var(mat) [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] [4,] [5,] summary(mat) V1 V2 V3 V4 V5 Min. :1.00 Min. :5.00 Min. : 9.00 Min. :13.00 Min. : st Qu.:1.75 1st Qu.:5.75 1st Qu.: st Qu.: st Qu.:17.75 Median :2.50 Median :6.50 Median :10.50 Median :14.50 Median :18.50 Mean :2.50 Mean :6.50 Mean :10.50 Mean :14.50 Mean : rd Qu.:3.25 3rd Qu.:7.25 3rd Qu.: rd Qu.: rd Qu.:19.25 Max. :4.00 Max. :8.00 Max. :12.00 Max. :16.00 Max. :20.00 mat <- matrix(1:20, ncol=5) mat [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] [4,] id <- mat[, 2] 5 id [1] FALSE TRUE TRUE TRUE mat[id, ] [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,]

33 矩陣的索引 33/60 (y <- array(1:15, dim=c(3, 5))) dim(y) [1] 3 5 x <- matrix(1:15, 3, 5) x [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] x[1] [1] 1 x[6] [1] 6 x <- matrix(1:15, 3, 5, byrow=true) x [,1] [,2] [,3] [,4] [,5] [1,] [2,] [3,] x[1] [1] 1 x[6] [1] 12 y[2, 4] [1] 11 y[1,] [1] y[,1] [1] y[2:3, ] [,1] [,2] [,3] [,4] [,5] [1,] [2,] y[-2,] [,1] [,2] [,3] [,4] [,5] [1,] [2,] y[,-2] [,1] [,2] [,3] [,4] [1,] [2,] [3,] dimnames(y) NULL rownames(y) NULL colnames(y) NULL

34 矩陣結合 (Forming Partitioned Matrices) cbind() and rbind() x <- c(1, 2, 3, 4, 5) y <- c(0.5, 0.4, 0.3, 0.2, 0.1) (z1 <- cbind(x,y)) x y [1,] [2,] [3,] [4,] [5,] (z2 <- rbind(x,y)) [,1] [,2] [,3] [,4] [,5] x y (A <- rbind(x,y)) [,1] [,2] [,3] [,4] [1,] [2,] [3,] [4,] [5,] [6,] [7,] (x <- matrix(1:20, ncol=4, nrow=5)) [,1] [,2] [,3] [,4] [1,] [2,] [3,] [4,] [5,] (y <- matrix(3:10, ncol=4)) [,1] [,2] [,3] [,4] [1,] [2,] (z <- matrix(rep(1:5, 2),nrow=5)) [,1] [,2] [1,] 1 1 [2,] 2 2 [3,] 3 3 [4,] 4 4 [5,] 5 5 (B <- cbind(x,z)) [,1] [,2] [,3] [,4] [,5] [,6] [1,] [2,] [3,] [4,] [5,] /60

35 常用矩陣運算 35/60

36 list(): 表列 36/60 List is an object consisting of an ordered collection of objects known as its components. A list could consist of a numeric vector, a logical value, a matrix, a complex vector, a character array, a function, and so on. 許多 R 統計函式回傳值皆是 list my.list <- list(name="george", wife="mary", no.children=3, child.ages=c(4,7,9)) my.list $name [1] "George" $wife [1] "Mary" $no.children [1] 3 $child.ages [1] Construct list: my.list <- list(name_1=object_1,,name_m=object_m) lst.abc <- list(list.a, list.b, list.c) see also:

37 表列 (List) 37/60 lst[1] vs lst[[1]] []: a general subscripting. [[]]: the operator used to select a single element. [1]: a sublist of the list consisting of the first entry. (name are included in the sublist). [[1]]: first object in the list, exclude name. my.list $name [1] "George" $wife [1] "Mary" $no.children [1] 3 $child.ages [1] my.list[[1]] # 傳回向量 [1] "George" my.list[[2]] [1] "Mary" my.list[[4]][1] [1] 4 getelement(mylist, "name") my.list$name # my.list[[1]] # my.list[[ name ]] [1] "George" my.list$wife # my.list[[2]] [1] "Mary" my.list$child.ages[1] # my.list[[4]][1] [1] 4 x <- "name" my.list[[x]] [1] "George" my.list[1] # 傳回 list $name [1] "George" my.list[2] $wife [1] "Mary"

38 課堂練習 6 38/60 my.list <- list(name=c("george", "John", "Tom"), wife=c("mary", "Sue", "Nico"), no.children=c(3, 2, 0), child.ages=list(c(4,7,9), c(2, 5), NA)) my.list$name [1] "George" "John" "Tom" my.list$wife [1] "Mary" "Sue" "Nico" my.list$no.children [1] my.list$name[3] [1] "Tom" my.list$name == "John" [1] FALSE TRUE FALSE my.list$child.ages [[1]] [1] [[2]] [1] 2 5 [[3]] [1] NA my.list$child.ages[2] [[1]] [1] 2 5 my.list$child.ages[[2]] [1] 2 5 my.list$child.ages[2][1] [[1]] [1] 2 5 my.list$child.ages[[2]][1] [1] 2 my.list$child.ages[[2]][2] [1] 5 length(my.list) [1] 4 my.list[[c(2, 3)]] [1] "Nico" my.list[c(2, 3)] $wife [1] "Mary" "Sue" "Nico" $no.children [1] 3 2 0

39 資料框 (Data Frame) 39/60 A data frame is a list with class data.frame. Regarded as a matrix with column possibly of differing modes and attributes. my.matrix <- matrix(1:15, ncol=3) my.matrix [,1] [,2] [,3] [1,] [2,] [3,] [4,] [5,] my.data <- data.frame(my.matrix) my.data X1 X2 X my.data[1, ] X1 X2 X my.data[2, 3] [1] 12 my.data$x1 [1] my.data[, "X1"] [1] my.data["x1"] X rownames(my.data) [1] "1" "2" "3" "4" "5" row.names(my.data) [1] "1" "2" "3" "4" "5" colnames(my.data) [1] "X1" "X2" "X3" names(my.data) [1] "X1" "X2" "X3"

40 資料框 (Data Frame) 40/60 rownames(my.data) <- c(paste("s", 1:5, sep=".")) colnames(my.data) <- c("a1", "A2", "A3") my.data A1 A2 A3 s s s s s subjects <-c ('Chinese', 'Math', 'English') scores <- c(50, 90, 61) pass <- scores = 60 student <- data.frame(subjects, scores, pass) student subjects scores pass 1 Chinese 50 FALSE 2 Math 90 TRUE 3 English 61 TRUE attach(my.data) A1 [1] A2 [1] A3 [1] detach() A1 Error: object "A1" not found student["2",] # use row names to extract records for no.2 subjects scores pass 2 Math 90 TRUE student[,"scores"] # use column names to extract values for "scores" [1]

41 列資料選取 41/60 index.1 <- iris[, "Species"] == "virginica" iris[index.1, ] Sepal.Length Sepal.Width Petal.Length Petal.Width Species virginica virginica... iris[species == "virginica",] Sepal.Length Sepal.Width Petal.Length Petal.Width Species virginica virginica... iris[!(species == "virginica"),] Sepal.Length Sepal.Width Petal.Length Petal.Width Species setosa setosa... m <- mean(iris$sepal.length) index.3 <- iris[, "Sepal.Length"] m iris[index.3, ] Sepal.Length Sepal.Width Petal.Length Petal.Width Species versicolor versicolor...

42 data.frame 欄位名稱 42/60 # data.frame 出來的資料欄位名稱, 會根據 input 所放是何種類別物件有關 x1 <- rnorm(5) x2 <- rnorm(5) x1 [1] class(x1) [1] "numeric" data.frame(var1=x1, Var2=x2) Var1 Var y1 <- data.frame(rnorm(5)) y2 <- data.frame(rnorm(5)) y1 rnorm class(y1) [1] "data.frame" data.frame(var1=y1, Var2=y2) rnorm.5. rnorm

43 課堂練習 43/60 class(iris) [1] "data.frame" head(iris, 3) Sepal.Length Sepal.Width Petal.Length Petal.Width Species setosa setosa setosa x1 <- iris[1] head(x1) Sepal.Length class(x1) [1] "data.frame" x2 <- iris[[1]] head(x2) [1] class(x2) [1] "numeric" x3 <- iris["sepal.length"] class(x3) [1] "data.frame" x4 <- iris[, "Sepal.Length"] class(x4) [1] "numeric" iris.copy <- iris iris.copy[3] <- NULL head(iris.copy, 3) Sepal.Length Sepal.Width Petal.Width Species setosa setosa setosa iris.copy <- iris iris.copy[[3]] <- NULL head(iris.copy, 3) Sepal.Length Sepal.Width Petal.Width Species setosa setosa setosa iris$species NULL iris$species [1] setosa setosa setosa... [145] virginica virginica virginica Levels: setosa versicolor virginica

44 課堂練習 : data.frame 的變數是一矩陣類別 44/60 iris.range <- aggregate(iris[, 1:4], by=list(iris[, 5]), range) str(iris.range) 'data.frame': 3 obs. of 5 variables: $ Group.1 : Factor w/ 3 levels "setosa","versicolor",..: $ Sepal.Length: num [1:3, 1:2] $ Sepal.Width : num [1:3, 1:2] $ Petal.Length: num [1:3, 1:2] $ Petal.Width : num [1:3, 1:2] iris.range Group.1 Sepal.Length.1 Sepal.Length.2 Sepal.Width.1 Sepal.Width.2 Petal.Length.1 Petal.Length.2 Petal.Width.1 Petal.Width.2 1 setosa versicolor virginica class(iris.range) [1] "data.frame" iris.range$sepal.length [,1] [,2] [1,] [2,] [3,] class(iris.range$sepal.length) [1] "matrix" dim(iris.range) [1] 3 5 iris.range[,1:3] Group.1 Sepal.Length.1 Sepal.Length.2 Sepal.Width.1 Sepal.Width.2 1 setosa versicolor virginica

45 日期 Dates 45/60 R 以 "Date" 類別表示 ( 不包括時間 ) 日期 : 年月日 Internally, Date objects are stored as the number of days since January 1, 1970, using negative numbers for earlier dates. The as.numeric function can be used to convert a Date object to its internal form. as.date(" ") [1] " " as.date("2019/02/17") [1] " " as.date(1000, origin = " ") [1] " " as.date("2/15/2011", format = "%m/%d/%y") [1] " " as.date("april 26, 1993", format = "%B %d, %Y") [1] " " as.date("22jun01", format = "%d%b%y") [1] " " seq(as.date(' '), by = 'days', length = 10) (lct <- Sys.getlocale("LC_TIME")) [1] "C" Sys.setlocale("LC_TIME", "C") [1] "C" [1] " " " " " " " " " " " " " " [8] " " " " " " seq(as.date(' '), to = as.date(' '), by='2 weeks') [1] " " " " " " " " " "

46 日期時間 46/60 R 的 " 時間 " 用 "POSIXct" 或 "POSIXlt" 類別表示, 內部時間是以 "1970 年 1 月 1 日 " 起至今的秒數表示 (UTC, Universal Time, Coordinated, 世界協調時間 ) (GMT, Greenwich Mean Time, 格林威治標準時間 ) Sys.time() [1] " :16:07 台北標準時間 " # extract date substr(as.character(sys.time()), 1, 10) [1] " " # extract time substr(as.character(sys.time()), 12, 19) [1] "21:16:07" date() [1] "Tue Oct 14 21:16: " now <- Sys.time() as.posixct(now) [1] " :46:44 CST" as.posixlt(now) [1] " :46:44 CST" class(now) [1] "POSIXct" "POSIXt" sec, min, hour, mday (# day number within the month), mon (#January=0), year (#+1900), wday (#day of the week starting at 0=sunday), yday (#day of the year after 1 january=0) my.date <- as.posixlt(sys.time()) my.date [1] " :18:31 台北標準時間 " my.date$sec [1] my.date$min [1] 18 my.date$hour [1] 21 my.date$mday [1] 14 my.date$mon [1] 9 my.date$year [1] 2028 my.date$wday [1] 2 my.date$yday [1] 287

47 POSIXlt / POSIXct 類別 47/60 Functions to convert between character representations and objects of classes "POSIXlt" and "POSIXct" representing calendar dates and times. Character input is first converted to class "POSIXlt" by strptime. Numeric input is first converted to "POSIXct". Any conversion that needs to go between the two date-time classes requires a time zone: conversion from "POSIXlt" to "POSIXct" will validate times in the selected time zone. as.posixct(" :59:59", format = "%Y-%m-%d %H:%M:%S", tz = "UTC") [1] " :59:59 UTC" as.posixlt(sys.time(), "GMT") [1] " :17:45 GMT"

48 strptime {base}: Date-time Conversion Functions to and from Character 48/60 x1 <- c(" ", " ", " ") strptime(x1, format="%y%m%d") [1] " CST" " CST" " CST" x2 <- c("27/02/2004", "27/02/2005", "14/01/2003") strptime(x2, format="%d/%m/%y") [1] " CST" " CST" " CST" x3 <- c("1jan1960", "2jan1960", "31mar1960", "30jul1960") strptime(x3, "%d%b%y") [1] " CST" " CST" " CST" " CDT" dates <- c("02/27/92", "02/27/92", "01/14/92", "02/28/92", "02/01/92") times <- c("23:03:20", "22:29:56", "01:03:30", "18:21:03", "16:56:26") x <- paste(dates, times) strptime(x, "%m/%d/%y %H:%M:%S") [1] " :03:20 CST" " :29:56 CST" [3] " :03:30 CST" " :21:03 CST" [5] " :56:26 CST" See also: Lubridate Package, Chron Packag

49 時間序列物件 : ts 49/60 ts(data = NA, start = 1, end = numeric(), frequency = 1, deltat = 1, ts.eps = getoption("ts.eps"), class =, names = ) as.ts(x,...) is.ts(x) ts(1:10, frequency = 4, start = c(1959, 2)) Qtr1 Qtr2 Qtr3 Qtr my.ts <- ts(1:10, frequency = 7, start = c(12, 2)) class(my.ts) [1] "ts" print(my.ts, calendar = TRUE) p1 p2 p3 p4 p5 p6 p gnp <- ts(cumsum(1+round(rnorm(100), 2)), start = c(1954, 7), frequency = 12) gnp Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec plot(gnp)

50 Multivariate ts 50/60 z <- ts(matrix(rnorm(300), 100, 3), start = c(1961, 1), frequency = 12) head(z, 3) Series 1 Series 2 Series 3 [1,] [2,] [3,] tail(z, 3) Series 1 Series 2 Series 3 [98,] [99,] [100,] class(z) [1] "mts" "ts" "matrix" plot(z) plot(z, plot.type = "single", lty = 1:3)

51 mode(object): 物件的模式 51/60 In R every "object" has a mode and a class. The former represents how an object is stored in memory (numeric, character, list and function) while the later represents its abstract type. Mode: "logical", "integer", "double", "complex", "raw", "character", "list", "expression", "name", "symbol" and "function". mode(object) Vector must have their values all of the same mode. Empty character string vector: character(0). Empty numeric vector: numeric(0). NOTE: Lists are of mode list. There are ordered sequences of objects which individually can be of any mode. (z <- 0:9) [1] mode(z) [1] "numeric" (digits <- as.character(z)) [1] "0" "1" "2" "3" "4" "5" "6" "7" "8" "9" mode(digits) [1] "character" (d <- as.integer(digits)) [1] mode(d) [1] "numeric" (x <- z[1:5]3) [1] FALSE FALSE FALSE FALSE TRUE mode(x) [1] "logical"

52 class(object): 物件的類別 52/60 For simple vector, mode=class: numeric, logical, character, list. matrix, array, factor, data.frame x1 <- 10 class(x1) [1] "numeric" (x2 <- seq(1, 10, 2)) [1] class(x2) [1] "numeric" my.f <- formula(iris$sepal.length ~ iris$sepal.width) class(my.f) [1] "formula" class(lm(my.f)) [1] "lm" class(aov(my.f)) [1] "aov" "lm" class(iris) [1] "data.frame" (iris.sub <- iris[5:10, 1:4]) Sepal.Length Sepal.Width Petal.Length Petal.Width class(iris.sub) [1] "data.frame" class(as.matrix(iris.sub)) [1] "matrix" as.list(iris.sub) $Sepal.Length [1] $Sepal.Width [1] $Petal.Length [1] $Petal.Width [1] class(as.list(iris.sub)) [1] "list"

53 class(object): 物件的類別 53/60 ex1 <- expression(1 + 0:9) # expression object ex1 expression(1 + 0:9) eval(ex1) [1] class(ex1) [1] "expression" hi <- function(){ + cat("hello world!\n") + } hi() hello world! class(hi) # function object [1] "function" # data frames are stored in memory as list but they are wrapped into data.frame objects. d <- data.frame(v1 = c(1,2)) class(d) [1] "data.frame" mode(d) [1] "list" typeof(d) [1] "list" (r.dates <- strptime(c("27/02/2004", "27/02/2005"), format="%d/%m/%y")) [1] " CST" " CST" class(r.dates) [1] "POSIXlt" "POSIXt" "There is a special object called NULL. It is used whenever there is a need to indicate or specify that an object is absent. It should not be confused with a vector or list of zero length. The NULL object has no type and no modifiable properties. There is only one NULL object in R, to which all instances refer. To test for NULL use is.null. You cannot set attributes on NULL. "

54 attributes(object): 物件的屬性 54/60 All objects except NULL can have one or more attributes attached to them. Select a specific attribute attr(object, name) Set a specific attribute attr(z, dim ) <- c(10, 10) x <- matrix(1:10, ncol=2) x [,1] [,2] [1,] 1 6 [2,] 2 7 [3,] 3 8 [4,] 4 9 [5,] 5 10 attributes(x) $dim [1] 5 2 attr(x, "dim") [1] 5 2 dim(x) [1] 5 2 x <- data.frame(matrix(1:10, ncol=2)) x X1 X attributes(x) $names [1] "X1" "X2" $row.names [1] $class [1] "data.frame" attr(x, "names") [1] "X1" "X2" names(x) [1] "X1" "X2" gender.f [1] 女女男女男男男男女女男 Levels: 女男 str(gender.f) Factor w/ 2 levels " 女 "," 男 ": class(gender.f) [1] "factor" attributes(gender.f) $levels [1] " 女 " " 男 " $class [1] "factor"

55 length(object): 物件的長度 55/60 beta <- c(1, 3, 5, 2, 4, 6, 11, NA, NA, 22) length(beta) [1] 10 length(beta[!is.na(beta)]) [1] 8 e <- numeric() # empty object e[3] <- 17 length(e) [1] 3 e [1] NA NA 17 (alpha <- numeric(10)) [1] length(alpha) [1] 10 alpha <- alpha[2*1:5] length(alpha) [1] 5 length(alpha) <- 3 alpha [1] mye <- expression(x, {y <- x^2; y+2}, x^y) length(mye) [1] 3 str(mye) expression(x, { y <- x^2 y + 2 }, x^y) myf <- formula(y ~ x1 + x2 + x3) length(myf) [1] 3 str(myf) Class 'formula' language y ~ x1 + x2 + x3..- attr(*, ".Environment")=<environment: R_GlobalEnv myf[1] `~`() myf[2] y() myf[3] (x1 + x2 + x3)() myf[4] Error in if (length(ans) == 0L... 需要 TRUE/FALSE 值的地方有缺值 mye[1] expression(x) mye[2] expression({ y <- x^2 y + 2 }) mye[3] expression(x^y) mye[4] expression(null)

56 str(object): 物件之結構 56/60 x <- 1:12 str(x) int [1:12] ch <- letters[1:12] str(ch) chr [1:12] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" str(iris) 'data.frame': 150 obs. of 5 variables: $ Sepal.Length: num $ Sepal.Width : num $ Petal.Length: num $ Petal.Width : num $ Species : Factor w/ 3 levels "setosa","versicolor",..: str(ls) # try str(str) function (name, pos = -1L, envir = as.environment(pos), all.names = FALSE, pattern, sorted = TRUE) str(str) function (object,...) myp <- plot(iris[,1], iris[,2]) str(myp) NULL (x <- as.date(" ")) [1] " " str(x) Date[1:1], format: " " (y <- strptime("27/02/2004", format="%d/%m/%y")) [1] " CST" str(y) POSIXlt[1:1], format: " "

57 str(object): 物件之結構 57/60 my.f <- iris$sepal.length ~ iris$sepal.width my.f iris$sepal.length ~ iris$sepal.width str(my.f) Class 'formula' language iris$sepal.length ~ iris$sepal.width..- attr(*, ".Environment")=<environment: R_GlobalEnv my.lm <- lm(my.f) my.lm Call: lm(formula = my.f) Coefficients: (Intercept) iris$sepal.width str(my.lm) List of 12 $ coefficients : Named num [1:2] $ qr :List of $ tol : num 1e my.lm$qr$tol [1] 1e-07 attr(my.lm$terms, "variables") list(iris$sepal.length, iris$sepal.width) $ terms :Classes 'terms', 'formula' language iris$sepal.length ~ iris$sepal.width....- attr(*, "variables")= language list(iris$sepal.length, iris$sepal.width)... - attr(*, "class")= chr "lm"

58 str(object): 物件之結構 58/60 my.lm.s <- summary(my.lm) my.lm.s Call: lm(formula = my.f) Residuals: Min 1Q Median 3Q Max mye <- expression(x, {y <- x^2; y+2}, x^y) mye expression(x, { y <- x^2 y + 2 }, x^y) Coefficients: Estimate Std. Error t value Pr( t ) (Intercept) <2e-16 *** iris$sepal.width Signif. codes: 0 *** ** 0.01 * Residual standard error: on 148 degrees of freedom Multiple R-squared: , Adjusted R-squared: F-statistic: on 1 and 148 DF, p-value: str(my.lm.s) List of 11 $ call : language lm(formula = my.f) $ terms :Classes 'terms', 'formula' language iris$sepal.length ~ iris$sepal.width....- attr(*, "variables")= language list(iris$sepal.length, iris$sepal.width)....- attr(*, "factors")= int [1:2, 1] attr(*, "class")= chr "summary.lm"

59 物件屬性強制轉換 (Coercing) 59/60 as.numeric(factor(c( a, b, c ))) as.numeric(c( a, b, c )) #don t work

60 資料類別 / 格式轉換 60/60 # converting rows of a matrix to elements of a list xy.matrix <- cbind(c(1, 2, 3), c(15, 16, 17)) xy.matrix [,1] [,2] [1,] 1 15 [2,] 2 16 [3,] 3 17 class(xy.matrix) [1] "matrix" xy.df <- data.frame(t(xy.location)) xy.df X1 X2 X class(xy.df) [1] "data.frame" xy.list <- as.list(xy.df) xy.list $X1 [1] 1 15 $X2 [1] 2 16 $X3 [1] 3 17 class(xy.list) [1] "list" Titanic,, Age = Child, Survived = No Sex Class Male Female 1st 0 0 2nd 0 0 3rd Crew 0 0,, Age = Adult, Survived = No Sex Class Male Female... class(titanic) [1] "table" data.frame(titanic) Class Sex Age Survived Freq 1 1st Male Child No 0 2 2nd Male Child No 0 3 3rd Male Child No 35 4 Crew Male Child No 0...