大数据分析 - R语言入门



本节旨在向用户介绍R编程语言。R可以从cran网站下载。对于Windows用户,安装RtoolsRStudio IDE非常有用。

R 的基本概念是作为用C、C++和Fortran等编译语言开发的其他软件的接口,并为用户提供一个交互式工具来分析数据。

导航到书籍zip文件bda/part2/R_introduction的文件夹,并打开R_introduction.Rproj文件。这将打开一个RStudio会话。然后打开01_vectors.R文件。逐行运行脚本并遵循代码中的注释。另一个有用的学习方法是直接键入代码,这将帮助你习惯R的语法。在R中,注释用#符号书写。

为了显示在书中运行R代码的结果,在代码评估后,R返回的结果会被注释掉。这样,你就可以将书中的代码复制粘贴到R中,并直接尝试其中的部分代码。

# 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"

让我们分析一下前面代码中发生的情况。我们可以看到,可以使用数字和字母创建向量。我们不需要事先告诉R我们想要什么类型的数据。最后,我们能够创建一个同时包含数字和字母的向量。向量mixed_vec已将数字强制转换为字符型,我们可以通过查看值是如何用引号打印出来的来看到这一点。

以下代码显示了class函数返回的不同向量的类型。通常使用class函数来“询问”对象,询问它的类是什么。

### 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也支持二维对象。在下面的代码中,有一些在R中使用的两种最流行的数据结构的示例:矩阵和数据框。

# 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,可以在同一个对象中使用不同的数据类型。通常,这就是数据在数据库、API中呈现的方式,部分数据是文本或字符向量,其他是数字。分析师的工作是确定要分配哪种统计数据类型,然后为其使用正确的R数据类型。在统计学中,我们通常认为变量属于以下类型:

  • 数值型
  • 名义型或类别型
  • 顺序型

在R中,向量可以是以下类:

  • 数值型 - 整数
  • 因子
  • 有序因子

R为每种统计类型的变量提供了一种数据类型。但是,有序因子很少使用,但可以通过函数factor或ordered创建。

下一节介绍索引的概念。这是一个相当常见的操作,它处理选择对象的部分并对其进行转换的问题。

# 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 
广告