R language [data management]

Catalog

One 、mode Access to type

  List related operations

 append operation

attach And detach

transform

  Two 、 Special values

  Check to see if there is NA

na.omit

  3、 ... and 、 Type conversion

  Four 、 Data sorting

sort()/order()

Sort a column in the table

  5、 ... and 、 Data set merging

1、 Direct horizontal consolidation

2. If you need to specify an index  

3. Vertical merger  

4. Example  

  6、 ... and 、 Get subset of data set

1、 Select the variable （ Column ）

2、 Remove variables （ Column ） 

 3. Selected Observations

  7、 ... and 、 Using functions subset() Select observation

  8、 ... and 、 Utility functions for processing data objects

  Nine 、 Vector grouping

Ten 、 Apply the function to each list element

  11、 ... and 、 Apply functions to rows or columns

  Twelve 、 Apply functions to group data

  13、 ... and 、 Apply functions to row groups

  fourteen 、 User defined functions

15、 ... and 、 loop  

---

One 、mode Access to type

stu_name <- "Richard"
yield <- 100
mode(stu_name)
mode(yield)

  List related operations

v <- c(1,2,3)
v
v <- c(v,4,5)
v
w <- c(6,7,8,9)
v <- c(v,w)
v[15] <- 0
v

 append operation

append You can append elements , In the specified after Then you can specify the location of the append

x <- 1:10
x
y <- append(x,100,after=5)
y
z <- append(x,100,after=0)
z

Here we can see that we can specify to add some two columns in the table , Averaging , The generated data will generate a new column of data .

mydata <- data.frame( x1 = c(1,2,3,4), x2 = c(5,6,7,8))
mydata
mydata$sumx <- mydata$x1+mydata$x2
mydata
mydata$meanx <- (mydata$x1+mydata$x2)/2
mydata

attach And detach

 attach And detach It's used in pairs .attach It can be connected to the table ,detach You can cancel the connection with the table

attach(mydata)
mydata$sumx <- x1+x2
mydata$meanx <- (x1+x2)/2
detach(mydata)
mydata <- transform(mydata, sumx=x1+x2, meanx=(x1+x2)/2) 

transform

Use transform To change the data frame

mydata <- transform(mydata, sumx=x1+x2, meanx=(x1+x2)/2)
mydata

  Two 、 Special values

From the following test , We can see if it's NULL Words ,R Language loss generates an object for us , But there is no storage space . If it is NA Words ,R Missing allocates a space to the object , But the content in this space is spatiotemporal .

NA Indicates a value that cannot be obtained

NULL Indicates a value that cannot be obtained because it does not exist

Inf infinite ,,1/0 What you get is Inf

NAN Can't express , It's not a number , Use Inf-Inf What you get is NAN

x <- NULL
y <- NA
length(x)
length(y)

  Check to see if there is NA

y <- c(1,2,3,NA)
is.na(y)

  If the data contains NA It cannot be calculated , But we can use na.rm=T Come and take our NA It's ignored

x <- y[1]+y[2]+y[3]+y[4]
x
z <- sum(y)
z
l <- sum(y, na.rm=T)
l

na.omit

  We can also use na.omit() To remove all of us including NA The observation of .

mydata <- read.table("/Users/yangkailiang/Documents/R/data/leadership.csv",
header=T, sep=",")
mydata
newdata <- na.omit(mydata)
newdata

  3、 ... and 、 Type conversion

Determine whether it is this type	Type conversion
is.character()	as.character()
is.vector()	as.vector()
is.matrix()	as.matrix()
is.data.frame()	as.data.frame()
is.factor()	as.factor()
is.logical()	as.logical()
is.numeric()	as.numeric()

a <- c(1,2,3)
a
is.numeric(a) 
is.vector(a) 
a <- as.character(a)
is.numeric(a)
is.vector(a) 
is.character(a) 

  Four 、 Data sorting

sort()/order()

We can see from the following test code , If used directly sort Its own words , The order of its own data will not change , Burning stone, if we send the sorted data to other variables to receive , The received data is in good order . The default order is ascending .

x <- c(12,4,7,11,2)
sort(x)
x
y <- sort(x)
y
z <- order(x)
x[z]

  Besides , We can also use order Make our data orderly . but order What is generated is the index value sorted by the size of each element in the array . So we need to use x[z] Make our data present an orderly state . As we can see from the figure below x The first element in should be placed in the fifth position , The second element should be in the second position , And so on .

 ​​​​​​​

Sort a column in the table

# According to our age Sort columns 
newdata <- mydata[order(mydata$age),] 
# According to our age Sort columns in descending order 
newdata2 <- mydata[order(mydata$age,decreasing=T),] 
# First according to gender Sort , Again according to age Sort in ascending order 
newdata3 <- mydata[order(mydata$gender,mydata$age),]

attach(mydata)
# First according to gender Sort , Then sort by age in descending order 
newdata4 <- mydata[order(gender,-age),]
newdata4
detach(mydata)

  5、 ... and 、 Data set merging

1、 Direct horizontal consolidation

Direct horizontal merge does not need to specify a public index

C = cbind(A,B)

2. If you need to specify an index  

C = merge(A,B, by=" Public index ")
C = merge(A,B, by="ID")
C = merge(A,B, by=c("ID","Country"))

3. Vertical merger  

If vertical consolidation is required ,A And B,A And B You need to have exactly the same variables （ The order can be different ）

4. Example  

Here are the data of our four tables

# Merge our two tables directly （ about ）
cdata <- cbind(data1,data3) 
data5 <- cbind(data1,data4) 

# Use merge Function words , The corresponding values will be merged automatically 
data6 <- merge(data1,data2)
data7 <- merge(data1,data4)

# according to kids Merge and sort columns 
data8 <- merge(data1,data4,by="kids")

  6、 ... and 、 Get subset of data set

1、 Select the variable （ Column ）

# From our mydata Select all rows , Take the first place 1,2,4 Column 
newdata <- mydata[,c(1,2,4)] 
# From our mydata Select all rows , take managerID,country,age Column 
newdata2 <- mydata[,c("managerID","country","age")]

2、 Remove variables （ Column ） 

mydata
# Method 1 
# To eliminate mydata The first row of planting , Second column , The fourth column 
newdata <- mydata[,c(-1,-2,-4)] 
newdata
# Method 2 
# To eliminate "managerID","country","age" These three columns , Will return a vector of boolean type 
#%in% For an operator 
delvar <- names(mydata) %in% c("managerID","country","age")
newdata2 <- mydata[!delvar] 

 3. Selected Observations

# Select one to three rows , All columns 
newdata <- mydata[1:3,] 
newdata

# The selected gender is M, And older than 30 The observation of 
newdata1 <- mydata[which(mydata$gender=="M" & mydata$age>30),] 
newdata1

attach(mydata)
# The selected gender is M, And older than 30 The observation of 
newdata2 <- mydata[which(gender=="M" & age>30),]
newdata2
detach(mydata)

  7、 ... and 、 Using functions subset() Select observation

# For us mydata Filter age ≥ in data frame 35 Or younger than 24 Observation of , And put it q1,q2,q3 The data of the column is displayed 
newdata <- subset(mydata, age>=35 | age < 24,select=c("q1","q2","q3"))
newdata
# take mydata The gender of the data frame is M, And older than 25, Will the gender List to q4 The columns are all displayed 
newdata2 <- subset(mydata, gender == "M" & age >25,select=gender:q4)
newdata2

  8、 ... and 、 Utility functions for processing data objects

 

length(newdata)
dim(newdata)
#str yes struct Abbreviation 
str(newdata)
class(newdata)
mode(newdata)
names(newdata)

head(newdata)
tail(newdata)
ls(newdata)

  Nine 、 Vector grouping

split() The return value is a list , Separate the data set according to a factor variable .

library(MASS)
groups<-split(Cars93$MPG.city,Cars93$Origin)
groups[[1]]
groups[[2]]

Ten 、 Apply the function to each list element

 lapply The return result is a list

sapply The return result is a vector

scores <- list(S1=numeric(0),S2=numeric(0),S3=numeric(0),S4=numeric(0))
scores$S1 <- c(89,85,85,62,93,77,85)
scores$S2 <- c(60,100,83,77,86)
scores$S3 <- c(95,86,91,82,63,67,97,64,55)
scores$S4 <- c(67,63,83,89)
lapply(scores,length) 
sapply(scores,mean)

 

 

  11、 ... and 、 Apply functions to rows or columns

apply()
results <- apply(matrix,1,function) # Process the rows of the matrix
results <- apply(matrix,2,function) # Process the columns of the matrix

For data frames , If the data type of each row is consistent , You can use apply To deal with it

If the column data types of the data frame are inconsistent , You need to use the above lapply And sapply To process

Mat_long <- matrix(c(-1.85,0.94,-0.54,-1.41,1.35,-1.71,-1.01,-0.16,-0.35,-3.72,1.63,-0.28,-0.28,2.45,-1.22),nrow=3,dimnames=list(c("Moe", "Larry","Curly"),c("trial1","trial2","trial3","trial4","trial")))
apply(Mat_long,1,mean)
apply(Mat_long,1,range)
apply(Mat_long,2,mean)

 

  Twelve 、 Apply functions to group data

tapply(vector,factor,function)

vector It's a vector ,factor Is a grouping factor ,function It's a function , That is to say, we need to be right vector The method of carrying out

wagedata <- read.csv("/Users/Documents/R/data/wagedata.csv")
attach(wagedata)
# First the edu according to female Grouping , Then calculate the average value respectively 
tapply(educ,female,mean)
detach(wagedata)
educ

 

 

  13、 ... and 、 Apply functions to row groups

by(dataframe,factor,function)

dataframe It's a vector ,factor Is a grouping factor ,function It's a function

wagedata <- read.csv("/Users/Documents/R/data/wagedata.csv")
by(wagedata,female,mean)

  fourteen 、 User defined functions

myfunction <- function(arg1,arg2,...){
statements
return(object)
}

# The following is the compilation of Pythagorean theorem 
# The square of a vector is the square of each component , Then we add the squares of each component and square it , We can get the length of our third side .
lengththm <- function(x) {
thirdl <- sqrt(sum(x^2))
return(thirdl)
}
a <- c(3,4)
lengththm(a)

 

15、 ... and 、 loop  

for (var in seq) {
statements
}
while (condition) {
statements
}

x <- c(2,5,10)
for (n in x)
print(n^2)