学习R之字符串和因子

1. 字符串

1.1 创建和输出字符串

字符向量可用 c 函数创建，可以用单引号或双引号把字符串引用起来，只要引号之间匹配即可，推荐使用双引号：

c("learn", "character", "and", "factor", "of", "r")
>> [1] "learn"     "character" "and"       "factor"    "of"        "r"

paste 函数能将不同字符串组合在一起，默认分隔符为空格，可以使用参数 sep 更改分隔符。

所有的字符串被组合后，可使用 collapse 参数把结果收缩成一个包含所有元素的字符串。

paste("red", "yellow")
>> [1] "red yellow"

paste(c("red", "yellow"), "lorry")
>> [1] "red lorry"    "yellow lorry"

paste(c("red", "yellow"), "lorry", sep = "-")
>> [1] "red-lorry"    "yellow-lorry"

paste("a", 1:5, sep = "")
>> [1] "a1" "a2" "a3" "a4" "a5"

paste("a", 1:5, sep = "", collapse = "+")
>> [1] "a1+a2+a3+a4+a5"

toString 函数是 paste 的变种，它在输出向量时，会使用逗号和空格分隔每个元素，同时可以限制输出的字符长度，如 width = 40 将输出 40 个字符：

x <- (1:15) ^ 2
toString(x)
>> [1] "1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 225"

toString(x, width = 40)
>> [1] "1, 4, 9, 16, 25, 36, 49, 64, 81, 100...."

1.2 格式化数字

format 函数可用于数字的格式化，包括指定小数的位数、输出的宽度、是否使用科学计数法等，传入的参数是数值型的向量，输出则是字符串向量：

pow <- 1:3
(powers_of_e <- exp(pow))
>> [1]  2.718282  7.389056 20.085537

format(powers_of_e)
>> [1] " 2.718282" " 7.389056" "20.085537"

format(powers_of_e, digits = 3)
>> [1] " 2.72" " 7.39" "20.09"

format(powers_of_e, digits = 3, width = 10) 
>> [1] "      2.72" "      7.39" "     20.09"

format(powers_of_e, digits = 3, scientific = TRUE)
>> [1] "2.72e+00" "7.39e+00" "2.01e+01"

sprint 函数通过指定占位符的方式对数字进行格式化输出，其中 %s 表示字符串，%d 表示整数，%f 和 %e 分别表示可指定小数位数的浮点数和科学计数法格式的浮点数。

sprintf("%s %d = %f", "Euler's constant to the power", 3, exp(3))
>> [1] "Euler's constant to the power 3 = 20.085537"

sprintf("To three decimal places, e ^ %d = %.4f", 3, exp(3))
>> [1] "To three decimal places, e ^ 3 = 20.0855"

sprintf("In scientific notation, e ^ %d = %e", 3, exp(3))
>> [1] "In scientific notation, e ^ 3 = 2.008554e+01"

1.3 更改大小写

使用 toupper 和 tolower 函数能把字符串中的字符全部转换为大写或小写：

toupper("I'm Shouting")
>> [1] "I'M SHOUTING"

tolower("I'm Whispering")
>> [1] "i'm whispering"

1.4 字符串操作

substr(x, start, stop) 函数是截取字符串最常用的函数，可以用来截取子字符串或者替换子字符串，其中 start 表示起始的索引，stop 表示结束的索引。

# 截取子字符串
substr("His name is Jack", 5, 8)
>> [1] "name"

# 替换子字符串
x <- "His name is Jack"
substr(x, 13, 16) <- "John"
x
>> [1] "His name is John"

strsplit 可以用来分割字符串，返回的是一个列表。

# 按空格分割字符串
x <- "His name is Jack"
strsplit(x, " ")
>> [[1]]
>> [1] "His"  "name" "is"   "Jack"

2. 因子

2.1 创建因子

因子是一个用于存储类别变量的特殊的变量类型，创建数据框时，R 会默认将类别变量转换为因子。

除了使用数据框在内部自动创建因子之外，也可以使用 factor 函数来创建因子。

heights <- data.frame(
       height_cm = c(153, 181, 150, 172, 165, 149, 174, 169, 198, 163),
       gender = c(
         "female", "male", "female", "male", "male",
         "female", "female", "male", "male", "female"
      )
)

class(heights$gender)
>> [1] "factor"

heights$gender
>>  [1] female male   female male   male   female female male   male   female
>> Levels: female male

gender_char <- c(
       "female", "male", "female", "male", "male",
       "female", "female", "male", "male", "female"
)
(gender_fac <- factor(gender_char))
>>  [1] female male   female male   male   female female male   male   female
>> Levels: female male

2.2 更改因子水平

可以通过指定 levels 参数来更改因子被创建时水平的先后顺序：

factor(gender_char, levels = c("male", "female"))
>>  [1] female male   female male   male   female female male   male   female
>> Levels: male female

2.3 去掉因子水平

在对数据进行处理的过程中，有时候需要去掉某些因子，如 getting_to_work 的数据框中剔除了 time_mins 字段为空的数据，但其对应的因子 bus 仍然被保留了。

getting_to_work <- data.frame(
       mode = c(
         "bike", "car", "bus", "car", "walk",
         "bike", "car", "bike", "car", "car"
       ),
      time_mins = c(25, 13, NA, 22, 65, 28, 15, 24, NA, 14)
)

getting_to_work$mode
>>  [1] bike car  bus  car  walk bike car  bike car  car 
>> Levels: bike bus car walk

# 剔除 time_mins 为空的
getting_to_work <- subset(getting_to_work, !is.na(time_mins))
unique(getting_to_work$mode)
>> [1] bike car  walk
>> Levels: bike bus car walk

如果要删除未使用的水平因子，我们可以使用 droplevels 函数，它接受因子或是数据框作为参数，丢弃输入因子中所有未使用的水平。

getting_to_work <- droplevels(getting_to_work)
levels(getting_to_work$mode)
>> [1] "bike" "car"  "walk"

2.4 有序因子

有些因子的水平在语义上大于或小于其他水平，这时可以使用有序因子，创建时指定参数 ordered = TRUE。

# 考试成绩 good < better < best
grade_choices <- c("good", "better", "best")
grade_values <- sample(grade_choices, 1000, replace = TRUE)
grade_fac <- factor(grade_values, grade_choices, ordered = TRUE)
head(grade_fac)
>> [1] best good good best best good
>> Levels: good < better < best
table(grade_fac)
>> grade_fac
>>   good better   best 
>>    363    293    344

2.5 连续变量转化为因子

cut 函数能将数值变量切成不同的组，然后返回一个因子。我们随机地生成 10000 名工人的年龄数据(从 16 到 66，使用 Beta 分布)，并将他们按每 10 年分组：

ages <- 16 + 50 * rbeta(10000, 2, 3)
grouped_ages <- cut(ages, seq.int(16, 66, 10))
head(grouped_ages)
>> [1] (36,46] (56,66] (46,56] (16,26] (26,36] (26,36]
>> Levels: (16,26] (26,36] (36,46] (46,56] (56,66]
table(grouped_ages)
>> grouped_ages
>> (16,26] (26,36] (36,46] (46,56] (56,66] 
>>    1788    3463    2960    1513     276

2.6 生成和合并因子

可以使用 gl 函数来生成因子，其中第一个参数为要生成的因子的水平数，第二个为每个水平需要重复的次数，第三个为因子的长度。

gl(3, 2)
>> [1] 1 1 2 2 3 3
>> Levels: 1 2 3

gl(3, 2, labels = c("A", "B", "C"))
>> [1] A A B B C C
>> Levels: A B C

gl(3, 1, 6, labels = c("A", "B", "C"))
>> [1] A B C A B C
>> Levels: A B C

如果有多个类别变量，有时把它们合并成一个单一的因子是有用的，其中每个水平由各个变量之间的交叉合并组成，可以使用 interaction 函数：

grade <- gl(3, 2, labels = c("A", "B", "C"))
gender <- gl(2, 1, 6, labels = c("female", "male"))
interaction(grade, gender)
>> [1] A.female A.male   B.female B.male   C.female C.male  
>> Levels: A.female B.female C.female A.male B.male C.male