R语言基础入门

2 * 3  #=> 6
sqrt(36)  #=> 6, square root
log10(100)  #=> 2, log base 10
10 / 3  #=> 3.3, 10 by 3
10 %/% 3  #=> 3, quotient of 10 by 3
10 %% 3  #=> 1, remainder of 10 by 3 余数　

a <- 10  # assign 10 to ‘a‘
a = 10  # same as above
10 -> a  # assign 10 to ‘a‘
10 = a  # Wrong!. This will try to assign `a` to 10.类

install.packages("car")  # install car package 下载离你所在地方最近的CRAN镜像的包 

library(car)  # initialize the pkg ‘car‘  初始化
require(car)  # another way to initialize 初始化
library()  # see list of all installed packages 列出所有以及下载的包
library(help=car)  # see info about ‘car‘ pkg 关于包的信息

help(merge)  # get help page for ‘merge‘
?merge  # lookup ‘merge‘ from installed pkgs 从以及下载的包中查找
??merge  # vague search 模糊查询
example(merge)  # show code examples 展示代码案例

getwd()  # gets the working directory 获取当前目录
setwd(dirname)  # set the working directory to dir name 设置工作目录 名称只能用 / 或者 \\ 分割 

myData <- read.table("c:/myInputData.txt", header = FALSE, sep="|", colClasses=c("integer","character","numeric") # import "|" separated .txt file

myData <- read.csv("c:/myInputData.csv", header=FALSE)  # import csv file 可以用colClass来人为设定数据框中每列的数据类型，这样处理效率更高，否则是自动识别数据类型的。
write.csv(rDataFrame, "c:/output.csv")  # export 　

a <- 10
b <- 20
ls()  # list objects in global env 列出全局环境中所有对象
rm(a)  # delete the object ‘a‘
rm(list = ls())  # caution: delete all objects in .GlobalEnv 删除全局环境中所有对象
gc()  # free system memory 释放系统内存

rm(list=ls())  # remove all objects in work space
env1 <- new.env()  # create a new environment
assign("a", 3, envir = env1)  # store a=3 inside env1
ls()  # returns objects in .GlobalEnv
ls(env1)  # returns objects in env1
get(‘a‘, envir=env1)  # retrieve value from env1

vec1 <- c(10, 20, 15, 40)  # numeric vector
vec2 <- c("a", "b", "c", NA)  # character vector
vec3 <- c(TRUE, FALSE, TRUE, TRUE)  # logical vector
vec4 <- gl(4, 1, 4, label = c("l1", "l2", "l3", "l4"))  # factor with 4 levels

length(vec1)  # 4
print(vec1[1])  # 10 从1开始索引 
print(vec1[1:3])  # 10, 20, 15 

numericVector <- numeric(100) # length 100 elements 

logic1 <- vec1 < 15  # create a logical vector, TRUE if value < 15
vec1[logic1]  # elements in TRUE positions will be included in subset 逻辑值为真的元素
vec1[1:2]  # returns elements in 1 & 2 positions.
vec1[c(1,3)]  # returns elements in 1 & 3 positions 位置1 3处元素

sort(vec1)  # ascending sort 升序
sort(vec1, decreasing = TRUE)  # Descending sort 降序

vec1[order(vec1)]  # ascending sort
vec1[rev(order(vec1))]  # descending sort

seq(1, 10, by = 2)  # diff between adj elements is 2 相邻元素差为2
seq(1, 10, length=25)  # length of the vector is 25 向量长度是25
rep(1, 5)  # repeat 1, five times. 重复1 五次
rep(1:3, 5)  # repeat 1:3, 5 times
rep(1:3, each=5)  # repeat 1 to 3, each 5 times.

vec2 <- c("a", "b", "c", NA)  # character vector
is.na(vec2)  # missing TRUE 返回一个缺失值处为真的逻辑向量
!is.na(vec2)  # missing FALSE
vec2[!is.na(vec2)]  # return non missing values from vec2 返回非缺失值

sample(x, size, replace = FALSE, prob = NULL)

set.seed(100)  # optional. set it to get same random samples. 设置随机数
sample(vec1)  # sample all elements randomly 把所有元素打乱顺序
sample(vec1, 3)  # sample 3 elem without replacement 抽取三个元素 3是抽样后得到的样本数量
sample(vec1, 10, replace=T)  # sample with replacement 意思是抽取的元素可以相同

#对数据框抽取
data=1:10
#抽取列
sample(x=data,size=1,replace=T)  
#抽取行
data[sample(nrow(data),2,replace=F),]

#按比例抽取
set.seed(1234)
#按7：3的比例产生了1和2
index <- sample(x = 2,size = nrow(data),replace=TRUE,prob = c(0.7,0.3))
traindata <- data[index == 1,]
testdata <- data[index == 2,]

#按照某个字段分层抽取
result<-data.frame()
data<-data.frame(a=c(1,2,3,4,6,7,8,5,11,12),b=c(2,3,4,5,65,4,6,6,8,9),s=c(‘A‘,‘B‘,‘A‘,‘A‘,‘B‘,‘A‘,‘B‘,‘A‘,‘B‘,‘B‘))
subdata<-split(data,data$s) #按照s变量中的类型分类成列表
for(i in 1:length(subdata)){
    sub<-subdata[[i]]
    x<-sub[sample(nrow(sub),2,replace=F),] #每个列表元素中抽取两行
    result<-rbind(result,x) #rbind函数 按照行叠加起来
}
result

myDf1 <- data.frame(vec1, vec2)  # make data frame with 2 columns
myDf2 <- data.frame(vec1, vec3, vec4)
myDf3 < data.frame(vec1, vec2, vec3)

library(datasets) # initialize
library(help=datasets) # display the datasets

#下面一些基本操作代码会经常用到 
class(airquality)  # get class
sapply(airquality, class)  # get class of all columns
str(airquality)  # structure
summary(airquality)  # summary of airquality
head(airquality)  # view the first 6 obs
fix(airquality)  # view spreadsheet like grid
rownames(airquality)  # row names
colnames(airquality)  # columns names
nrow(airquality)  # number of rows
ncol(airquality)  # number of columns

# cbind  rbind 为数据框添加行和列
cbind(myDf1, myDf2)  # columns append DFs with same no. rows
rbind(myDf1, myDf1)  # row append DFs with same no. columns

myDf1$vec1  # vec1 column
myDf1[, 1]  # df[row.num, col.num]
myDf1[, c(1,2)]  # columns 1 and 2
myDf1[c(1:5), c(2)]  # first 5 rows in column 2

#也可以使用 subset() 和 which() 函数 which() 返回符合条件的行或列向量
subset(airquality, Day == 1, select = -Temp)  # 选择 Day=1 剔除 ‘Temp‘
airquality[which(airquality$Day==1), -c(4)]  # 作用同上

set.seed(100)
trainIndex <- sample(c(1:nrow(airquality)), size=nrow(airquality)*0.7, replace=F)  # get test sample indices 0.7比例 灵活性
airquality[trainIndex, ]  # training data
airquality[-trainIndex, ]  # test data

paste("a", "b")  # "a b"
paste0("a", "b")  # concatenate without space, "ab"
paste("a", "b", sep="")  # same as paste0
paste(c(1:4), c(5:8), sep="")  # "15" "26" "37" "48"
paste(c(1:4), c(5:8), sep="", collapse="")  # "15263748"
paste0(c("var"), c(1:5))  # "var1" "var2" "var3" "var4" "var5"
paste0(c("var", "pred"), c(1:3))  # "var1" "pred2" "var3"
paste0(c("var", "pred"), rep(1:3, each=2))  # "var1" "pred1" "var2" "pred2" "var3" "pred3

dateString <- "15/06/2014"
myDate <- as.Date(dateString, format="%d/%m/%Y")
class(myDate)  # "Date"
myPOSIXctDate <- as.POSIXct(myDate)  # convert to POSIXct

attributes(myPOSIXltDate)  # best
unclass(POSIXltDate)  # works!
names(myPOSIXltDate)  # doesn‘t work on a POSIXlt object
unlist(myPOSIXltDate)  # works!

object.size(myDate)  # 216 bytes
object.size(myPOSIXltDate)  # 1816 bytes
object.size(myPOSIXctDate)  # 520 bytes

test <- c(1,2,3,2,1,1,5,6,4,5,1)
test1 <- table(test)
test1
test
1 2 3 4 5 6   #test中的元素
4 2 1 1 2 1   #各个元素出现的频率
#提取table()中的元素
names(test1)
[1] "1" "2" "3" "4" "5" "6"
#提取table()中的频率
as.numeric(test1)
[1] 4 2 1 1 2 1
#3.进阶，提取table()中指定频率的数据，先用as.data.frame()转换
test2 <- as.data.frame(test1)
test2
  test Freq
1    1    4
2    2    2
3    3    1
4    4    1
5    5    2
6    6    1
test2[which(test2$Freq==1),]  #提取出现频率为1的元素
  test Freq
3    3    1
4    4    1
6    6    1

# Create a list containing strings, numbers, vectors and a logical values.
> list_data <- list("Red", "Green", c(21,32,11), TRUE, 51.23, 119.1)
> print(list_data)
> print(list_data)
[[1]]
[1] "Red"
[[2]]
[1] "Green"
[[3]]
[1] 21 32 11
[[4]]
[1] TRUE
[[5]]
[1] 51.23
[[6]]
[1] 119.1

# Give names to the elements in the list.
 names(list_data) <- c("1st Quarter", "A_Matrix", "A Inner list")

print(list_data[1])#使用元素索引访问
print(list_data$A_Matrix)#使用名称访问

list_data[3] <- "updated element"#更新
list_data[4] <- NULL #删除
# Create two lists.
list1 <- list(1,2,3)
list2 <- list("Sun","Mon","Tue")
# Merge the two lists.
 merged.list <- c(list1,list2)

# Convert the lists to vectors.
v1 <- unlist(list1)

if(checkConditionIfTrue) {
  ....statements..
  ....statements..
} else {   # place the ‘else‘ in same line as ‘}‘
  ....statements..
  ....statements..
} 

for(counterVar in c(1:n)){
  .... statements..
}

tryCatch({1 <- 1; print("Lets create an error")}, # First block
error=function(err){print(err); print("Error Line")},  # Second Block(optional)
finally = {print("finally print this")})# Third Block(optional)
#=> [1] "Lets create an error"
#=> <simpleError in 1 <- 1: invalid(do_set) left-hand side to assignment>
#=> [1] "Error Line"
#=> [1] "finally print this"