大資料分析 - R語言入門

本節旨在向用戶介紹R程式語言。R可以從cran網站下載。對於Windows使用者，安裝Rtools和RStudio IDE非常有用。

R 的基本概念是作為用C、C++和Fortran等編譯語言開發的其他軟體的介面，併為使用者提供一個互動式工具來分析資料。

導航到書籍zip檔案bda/part2/R_introduction的資料夾，並開啟R_introduction.Rproj檔案。這將開啟一個RStudio會話。然後開啟01_vectors.R檔案。逐行執行指令碼並遵循程式碼中的註釋。另一個有用的學習方法是直接鍵入程式碼，這將幫助你習慣R的語法。在R中，註釋用#符號書寫。

為了顯示在書中執行R程式碼的結果，在程式碼評估後，R返回的結果會被註釋掉。這樣，你就可以將書中的程式碼複製貼上到R中，並直接嘗試其中的部分程式碼。

# Create a vector of numbers 
numbers = c(1, 2, 3, 4, 5) 
print(numbers) 

# [1] 1 2 3 4 5  
# Create a vector of letters 
ltrs = c('a', 'b', 'c', 'd', 'e') 
# [1] "a" "b" "c" "d" "e"  

# Concatenate both  
mixed_vec = c(numbers, ltrs) 
print(mixed_vec) 
# [1] "1" "2" "3" "4" "5" "a" "b" "c" "d" "e"

讓我們分析一下前面程式碼中發生的情況。我們可以看到，可以使用數字和字母建立向量。我們不需要事先告訴R我們想要什麼型別的資料。最後，我們能夠建立一個同時包含數字和字母的向量。向量mixed_vec已將數字強制轉換為字元型，我們可以透過檢視值是如何用引號打印出來的來看到這一點。

以下程式碼顯示了class函式返回的不同向量的型別。通常使用class函式來“詢問”物件，詢問它的類是什麼。

### Evaluate the data types using class

### One dimensional objects 
# Integer vector 
num = 1:10 
class(num) 
# [1] "integer"  

# Numeric vector, it has a float, 10.5 
num = c(1:10, 10.5) 
class(num) 
# [1] "numeric"  

# Character vector 
ltrs = letters[1:10] 
class(ltrs) 
# [1] "character"  

# Factor vector 
fac = as.factor(ltrs) 
class(fac) 
# [1] "factor"

R也支援二維物件。在下面的程式碼中，有一些在R中使用的兩種最流行的資料結構的示例：矩陣和資料框。

# Matrix
M = matrix(1:12, ncol = 4) 
#      [,1] [,2] [,3] [,4] 
# [1,]    1    4    7   10 
# [2,]    2    5    8   11 
# [3,]    3    6    9   12 
lM = matrix(letters[1:12], ncol = 4) 
#     [,1] [,2] [,3] [,4] 
# [1,] "a"  "d"  "g"  "j"  
# [2,] "b"  "e"  "h"  "k"  
# [3,] "c"  "f"  "i"  "l"   

# Coerces the numbers to character 
# cbind concatenates two matrices (or vectors) in one matrix 
cbind(M, lM) 
#     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] 
# [1,] "1"  "4"  "7"  "10" "a"  "d"  "g"  "j"  
# [2,] "2"  "5"  "8"  "11" "b"  "e"  "h"  "k"  
# [3,] "3"  "6"  "9"  "12" "c"  "f"  "i"  "l"   

class(M) 
# [1] "matrix" 
class(lM) 
# [1] "matrix"  

# data.frame 
# One of the main objects of R, handles different data types in the same object.  
# It is possible to have numeric, character and factor vectors in the same data.frame  

df = data.frame(n = 1:5, l = letters[1:5]) 
df 
#   n l 
# 1 1 a 
# 2 2 b 
# 3 3 c 
# 4 4 d 
# 5 5 e 

如前面的示例所示，可以在同一個物件中使用不同的資料型別。通常，這就是資料在資料庫、API中呈現的方式，部分資料是文字或字元向量，其他是數字。分析師的工作是確定要分配哪種統計資料型別，然後為其使用正確的R資料型別。在統計學中，我們通常認為變數屬於以下型別：

• 數值型
• 名義型或類別型
• 順序型

在R中，向量可以是以下類：

• 數值型 - 整數
• 因子
• 有序因子

R為每種統計型別的變數提供了一種資料型別。但是，有序因子很少使用，但可以透過函式factor或ordered建立。

下一節介紹索引的概念。這是一個相當常見的操作，它處理選擇物件的部分並對其進行轉換的問題。

# Let's create a data.frame
df = data.frame(numbers = 1:26, letters) 
head(df) 
#      numbers  letters 
# 1       1       a 
# 2       2       b 
# 3       3       c 
# 4       4       d 
# 5       5       e 
# 6       6       f 

# str gives the structure of a data.frame, it’s a good summary to inspect an object 
str(df) 
#   'data.frame': 26 obs. of  2 variables: 
#   $ numbers: int  1 2 3 4 5 6 7 8 9 10 ... 
#   $ letters: Factor w/ 26 levels "a","b","c","d",..: 1 2 3 4 5 6 7 8 9 10 ...  

# The latter shows the letters character vector was coerced as a factor. 
# This can be explained by the stringsAsFactors = TRUE argumnet in data.frame 
# read ?data.frame for more information  

class(df) 
# [1] "data.frame"  

### Indexing
# Get the first row 
df[1, ] 
#     numbers  letters 
# 1       1       a  

# Used for programming normally - returns the output as a list 
df[1, , drop = TRUE] 
# $numbers 
# [1] 1 
#  
# $letters 
# [1] a 
# Levels: a b c d e f g h i j k l m n o p q r s t u v w x y z  

# Get several rows of the data.frame 
df[5:7, ] 
#      numbers  letters 
# 5       5       e 
# 6       6       f 
# 7       7       g  

### Add one column that mixes the numeric column with the factor column 
df$mixed = paste(df$numbers, df$letters, sep = ’’)  

str(df) 
# 'data.frame': 26 obs. of  3 variables: 
# $ numbers: int  1 2 3 4 5 6 7 8 9 10 ...
# $ letters: Factor w/ 26 levels "a","b","c","d",..: 1 2 3 4 5 6 7 8 9 10 ... 
# $ mixed  : chr  "1a" "2b" "3c" "4d" ...  

### Get columns 
# Get the first column 
df[, 1]  
# It returns a one dimensional vector with that column  

# Get two columns 
df2 = df[, 1:2] 
head(df2)  

#      numbers  letters 
# 1       1       a 
# 2       2       b 
# 3       3       c 
# 4       4       d 
# 5       5       e 
# 6       6       f  

# Get the first and third columns 
df3 = df[, c(1, 3)] 
df3[1:3, ]  

#      numbers  mixed 
# 1       1     1a
# 2       2     2b 
# 3       3     3c  

### Index columns from their names 
names(df) 
# [1] "numbers" "letters" "mixed"   
# This is the best practice in programming, as many times indeces change, but 
variable names don’t 
# We create a variable with the names we want to subset 
keep_vars = c("numbers", "mixed") 
df4 = df[, keep_vars]  

head(df4) 
#      numbers  mixed 
# 1       1     1a 
# 2       2     2b 
# 3       3     3c 
# 4       4     4d 
# 5       5     5e 
# 6       6     6f  

### subset rows and columns 
# Keep the first five rows 
df5 = df[1:5, keep_vars] 
df5 

#      numbers  mixed 
# 1       1     1a 
# 2       2     2b
# 3       3     3c 
# 4       4     4d 
# 5       5     5e  

# subset rows using a logical condition 
df6 = df[df$numbers < 10, keep_vars] 
df6 

#      numbers  mixed 
# 1       1     1a 
# 2       2     2b 
# 3       3     3c 
# 4       4     4d 
# 5       5     5e 
# 6       6     6f 
# 7       7     7g 
# 8       8     8h 
# 9       9     9i