R语言向量与因子

使用combine创建向量

# 字符型向量
name <- c("张三","李四")
# 数值型向量
num <- c(1,2,3.4,5)
# 逻辑型向量
b <- c(F,T,FALSE,TRUE)

# 不能有混合类型
a <- c(1,2,T,F) # 1 2 1 0
# 不存在子向量
d <- c(1,c(2,3),c(4,5)) # 1 2 3 4 5

# 创建固定长度向量
x1 <- vector("numeric",3) # 0 0 0
x2 <- numeric(3) # 0 0 0
x3 <- character(3) # "" "" ""
x4 <- logical(3) # FALSE FALSE FALSE
x5 <- vector(length = 3) # FALSE FALSE FALSE

使用seq创建向量

seq(from = 1,to = 10,by=2) # 1 3 5 7 9
seq(from = 10,to = 1,by=-2) # 10  8  6  4  2
seq(from = 1,to = 10,len=3) # 1.0  5.5 10.0
# 特别的，步长为1
1:5 # 1 2 3 4 5
pi:1 # 3.141593 2.141593 1.141593
1:5-1 # 0 1 2 3 4
1:(5-1) # 1 2 3 4

使用sample创建向量

sample(5) # 2 4 5 1 3
sample(c('a','b','c','d')) # "d" "c" "b" "a"
set.seed(2020) # 设置随机数种子
sample(5) # 5 2 4 3 1
sample(1:5,3) # 1 5 2 随机选三个

# 有放回的抽样
re_sample = sample(1:100,100,replace = TRUE)
unique_re_sample  = unique(re_sample)
length(unique_re_sample)

访问向量子集

正整数下标

score <- c(95,96,85,98,88,90)
score[c(3,5)] # 85 88
score[-c(3,5)] # 95 96 98 90 反向取出
score[c(3,5)] - 90 # -5 -2
score[c(3,5)] <- score[c(3,5)] +6
score #  95 96 91 98 94 90

注意下标的特殊用法

score[] <- mean(score) # 每一个元素获得平均分
score  # 94 94 94 94 94 94
score <- mean(score) # 一个数值平均分
score # 94
# 下标可以重复，顺序可以改变
name <- c("张三","李四","王五")
name[c(1,1,3,2)] # "张三" "张三" "王五" "李四"

负整数下标

score <- c(95,96,85,98,88,90)
score[-c(3,5)] # 95 96 98 90
idx <- which(score<90) #  3 5 小于90的下标
score[-idx] # 95 96 98 90

逻辑下标

score <- c(95,96,85,98,88,90)
name <- c("张三","李四","王五","刘备","曹操","张飞")
score < 90 #  FALSE FALSE  TRUE FALSE  TRUE FALSE
score[score<90] # 85 88
name[score<90] #  "王五" "曹操" 小于90的姓名

通过元素的名称访问子集

score <- c(95,96,85,98,88,90)
xm <- c("张三","李四","王五","刘备","曹操","张飞")
names(score)<- xm
score
# 张三 李四 王五 刘备 曹操 张飞 
# 95   96   85   98   88   90 
score[c("刘备","张飞")]
# 刘备 张飞 
# 98   90 

向量的基本操作

向量排序

v1 <- c(a=5,b=10,c=12,d=6)
sort(v1)
# a  d  b  c 
# 5  6 10 12 
order(v1,decreasing = TRUE) # 3 2 4 1 下标排序
v1[order(v1,decreasing = TRUE)]
# c  b  d  a 
# 12 10  6  5 
score <- c(95,96,85,98,88,90)
rev(score) # 90 88 98 85 96 95
score[length(score)] # 90 取最后一个元素
tail(score,n=1) # 90 取最后一个元素
rev(tail(score,n=3)) # 90 88 98 倒数3个元素 

向量的运算

p0 <- c(0,0)
p1 <- c(1,2)
p2 <- c(2,1)
# 求和
p3 <- p1+p2 # 3 3
# 数乘
p4 <- 1.5*p3 # 4.5 4.5

p1_on_p2 <- sum(p1*p2)/
        sum(p1*p2)*p2 # 2 1 计算投影向量  

因子的创建

gender <- c("male","male","female","female")
typeof(gender)
# [1] "character"
gender
# [1] "male"   "male"   "female" "female"
gender <- factor(gender)
typeof(gender)
# [1] "integer"
gender
# [1] male   male   female female
# Levels: female male

因子的操作

gender <- c("male","male","female","female")
gender <- factor(gender)
gender[c(1,2:3)]
# [1] male   male   female
# Levels: female male

nlevels(gender) # 2
levels(gender)# [1] "female" "male"  

gender[1]<-"female" # 此时只能赋值"female" "male"  
gender
# [1] female male   female female
# Levels: female male

定义因子

gender <- c("male","male","female","female")
gender <- factor(gender,levels = c("male","female","shemale"))
gender[1] <- "shemale"
gender
# [1] shemale male    female  female 
# Levels: male female shemale

因子的本质

gender <- c("male","male","female","female")
gender <- factor(gender)
as.numeric(gender)
# [1] 2 2 1 1
as.character(gender)
# [1] "male"   "male"   "female" "female"

number_factor <- c(10,20,40,20,30,10,20)
number_factor <- factor(number_factor)
as.numeric(number_factor) # [1] 1 2 4 2 3 1 2
# 因子正确求平均值
mean(as.numeric(as.character(number_factor))) # [1] 21.42857

mean(as.numeric(levels(number_factor)[number_factor])) # [1] 21.42857

创建有序因子

score <- factor(c('优','良','中','差','优','良','中'))
# error score[1]<score[2]
score <- factor(c('优','良','中','差','优','良','中'),ordered = TRUE)
score[1] <score[2] # TEUE 默认按字母排序 y l z c

score <- factor(c('优','良','中','差','优','良','中')
                ,ordered = TRUE
                ,levels = c('差','中','良','优'))

score[1] <score[2] # FALSE 中 > 差

根据已有的数据进行分箱

# 将百分制转为5分制
score <- c(94,87,92,91,85,92)
score_factor_5 <- cut(score,
                      breaks = c(0,(6:10)*10),
                      include.lowest = TRUE,
                      right = FALSE,
                      ordered_result = TRUE,
                      labels = c('不及格','及格','中','良','优'))
score_factor_5
# [1] 优 良 优 优 良 优
# Levels: 不及格 < 及格 < 中 < 良 < 优