• setwd("path")
• getwd()#pwd for working dir
• read.csv("xxx.csv")
• dir() #ls
• ls() #list all obj
• source("codepath") #load code
• code fucntion name to show code

objects:

• character
• numeric
• integer
• complex
• logical

datastruct

index starts from 1

vector & list: support same&diff class

list like excel table

Inf: special number,

• NaN:numrecial missing value

• NA:others

  is.na(x) is.nan(x)

attributes()

c(T,F) #create vectors list(1,"a")#create list

as.logical()

matrix

m <- matrix(nrow = 2,ncol = 3)
m2 <- matrix1:6,nrow = 2,ncol = 3)
dim(m)
dim

#transform
m <- 1:10
dim(m) <- c(2,5)

     [,1] [,2] [,3] [,4] [,5]
[1,]    1    3    5    7    9
[2,]    2    4    6    8   10

#binding col and row
cbind(x,y,z)
 x y  z
[1,] 1 7 -3
[2,] 2 8 -2
[3,] 3 9 -1
rbind(x,y,z)
[,1] [,2] [,3]
x    1    2    3
y    7    8    9
z   -3   -2   -1

use drop arg to get subsetting

factors

分类数据（labels），order or unordered

x <- factor(c("rrr","ffff"."rrr"),levels = c("ffff","rrr"))

data frames

matrix thar support diff type;col obj must be same type

read.table() read.csv()

data.matrix(x)

> x <- data.frame(foo = 7:9,bar = c(T,F, F))
> x
  foo   bar
1   7  TRUE
2   8 FALSE
3   9 FALSE

diff list : ()

• you can't use the same name for two different variables (as list)
• all elements of a data frame are vectors
• all elements of a data frame have an equal length.

  names attribute

vector list matrix

dimnames(matrix)

IO

opt read.table

for large dataset read help page to opt usage of read.table()

comment.char = ""

colClasses try to figure out data type every col

intial <- read.table("datatable.txt", nrows = 100) #read 100 rows data;help r to init memory
classes <- sapply(initial,calss)
tabAll <- read.table("datatable.txt", colClasses = classes)

almost 2*data_requirs memory

dput dget & dmup source

to out

• gzfile
• bzfile
• file
• url

readLines

[] [[]] $ subset

[index] : returns obj as parent ;;index as colum-index

[[index]] : returns item ---- list(sequence)

[[$name_index]] or x$name)index : returns item by name

for list

x[c(3,8)] #return 3th seq and 8th seq

x[[c(3,8)]] #return item on 3th rou/seq and 8th col

for matrix

x[1,2] reutrns item-vector in matrix

x[1,2, drop = FLASE] reutrns matrix

numerical index and logical index

partial matching

x <- list(aaa= 1:5)
x$a
x[["a", exact = F]]

多个匹配return NULL

complete.cases(x,y)

logical or

%*% true matrix multiplication

condition control

if - else

y <- if(x>3) {
    10
} else {3}

for

for(i in seq_along(array)) { }

for(i in seq_len(nrou(x))) {
    for(j in seq_len(ncol(x))) {}
     }

repeat

repeat {
    ...
    break
}

break next return

function returns last expression value

function

a functuon can be defined in function

name <- function(x) {}

last value as return value

arument matching

by name arguments doesn't need to be ordered

1. name match
2. partial match
3. positional match

lazy evaluation

eval argument only when it is needed

... argument

partial match and positional match can't used after ...

>args(function_name)

environment order

search()

函数内修改全局变量，仅在函数作用域内生效

Date Time

time POSIXct

x <- as.Date("1970-01-01") # value 0
# as.POSIClt
y <- Sys.time()

use function strptime