Big Data Analytics 简明教程
Big Data Analytics - Introduction to R
本节专门向用户介绍 R 编程语言。可从 cran website 下载 R。对于 Windows 用户, install rtools 和 rstudio IDE 很有用。
R 背后的总体概念是充当已编译语言(例如 C、C++ 和 Fortran)编写的其他软件的界面,并为用户提供一个交互式工具来分析数据。
导航到图书 zip 文件的文件夹 bda/part2/R_introduction 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。为了学习,另一个有用的选择就是键入代码,这将帮助您习惯 R 语法。在 R 中,注释是用 # 符号编写的。
为了在图书中显示运行 R 代码的结果,在评估代码后,将注释返回的结果 R。通过这种方式,您可以在书中复制粘贴代码并直接在 R 中尝试其部分。
# 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"
让我们分析一下之前代码中发生的情况。我们可以看到可以使用数字和字母创建向量。我们无需事先告诉 R 我们想要什么类型的数据类型。最后,我们能够创建带有数字和字母的向量。混合向量已将数字强制转换为字符,我们可以通过可视化如何将值打印在引号内来看到这一点。
以下代码显示了函数类返回的不同向量的类型。通常使用类函数来“询问”对象,询问他的类别是什么。
### 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。在以下代码中,列出了 R 中使用最流行的两个数据结构的示例:矩阵和数据帧。
# 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
如前一个示例所示,可以在同一对象中使用不同的数据类型。通常,这是如何在数据库中呈现数据,API 数据的一部分是文本或字符向量,另一部分是数字。分析人员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。在统计中,我们通常认为变量为以下类型 −
-
Numeric
-
Nominal or categorical
-
Ordinal
在 R 中,向量可以属于以下类 −
-
Numeric - Integer
-
Factor
-
Ordered Factor
R 为每种变量的统计类型提供数据类型。然而,有序因子很少使用,但可以通过因子函数或 ordered 函数创建。
以下部分讨论了索引的概念。这是一个非常常见的操作,并且处理选择对象的各个部分并对它们进行转换的问题。
# 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