Apply, TApply, LApply, Vapply, Ftable, xtab and aggregate functions are very important for data transformation. These are basic data processing functions. Script used in above lecture is mentioned below:

#Yogesh Mehla
#yogesh.mehla@gmail.com
#+91-9023262520

#apply, lapply, sapply, tapply, aggregate, xtab 
#generating matrix(cross tab reports)

CricData<-read.csv('G:/R/Rpractice.csv')
CricData

for(i in 1:length(CricData))
{
  print(class(CricData[,i]))
}

CricData[,2]<-as.character(CricData[,2])
CricData[,3]<-as.character(CricData[,3])
CricData[,4]<-as.character(CricData[,4])
CricData[,7]<-as.character(CricData[,7])

for(i in 1:length(CricData))
{
  print(class(CricData[,i]))
}

for(i in 1:length(CricData))
{
  if(class(CricData[,i])=="integer" || class(CricData[,i])=="numeric")
  {
    print(paste("mean of column ", colnames(CricData[,i]), " is " ,mean(CricData[,i])))
  }
  else
  {
    print(paste("Column ",  colnames(CricData[,i]), " is of type",class(CricData[,i]) ))
  }
}

#apply implement function on either row or column depending upon MARGIN parameter
apply(mtcars, 2, class)

apply(mtcars,2, mean)

apply(mtcars,2, name <- function(x) {
  print(paste("mean is ",mean(x)))
})

apply(CricData,2, class)

apply(CricData[5:6],2,class)

#sapply it by default implement function on column, no MARGIN parameter is available for sapply
sapply(CricData,name <- function(x) {
  print(paste("mean is ",mean(x)))
})

sapply(CricData,name <- function(x) {
  if(class(x)=="integer" || class(x)=="numeric")
  {
    print(paste("mean of column ", colnames(x), " is " ,mean(x)))
  }
  else
  {
    print(paste("Column ",  colnames(x), " is of type",class(x) ))
  }
})

apply(CricData[5:6],2,mean)

fix(CricData)

apply(CricData[5:6],2,mean,na.rm=TRUE)

#lapply it by default implement function on column, no MARGIN parameter is available for lapply
#lapply always return data in form of list
lapply(CricData[5:6],mean,na.rm=TRUE)

#unlist of returned output
unlist(lapply(CricData[5:6],mean,na.rm=TRUE))

CricData$Franchise<-as.character(CricData$Franchise)

unique(CricData$Franchise)

OutputList<-lapply(unique(CricData$Franchise), name <- function(x) {
  CricData[CricData$Franchise==x,]
})
OutputList[1]

UniqueFranchise<-unique(CricData$Franchise)
OutputList<-lapply( setNames(UniqueFranchise,UniqueFranchise), function(x) {
  CricData[CricData$Franchise==x,]
})
OutputList$Jaipur

#split function do similar task
OutputSplit<-split(CricData,CricData$Franchise)
OutputSplit$Delhi

class(OutputList)
class(OutputSplit)

unsplit(OutputSplit,CricData$Franchise)
unsplit(OutputList,CricData$Franchise)

#vapply is similar to sapply but you need to specify return type
vapply(CricData[5:6], mean ,numeric(1))

#tapply group values one base of INDEX parameter. tapply returns array
tapply(CricData$Price, CricData$Specialty,sum)

tapply(CricData$Price, list(CricData$Franchise, CricData$Specialty),sum)

tapply(CricData$Price, list(Franchise=CricData$Franchise, Specialty=CricData$Specialty),sum)

CricData[order(CricData$Franchise),]

CricData[order(CricData$Franchise,CricData$Specialty),]

CricData[order(-as.numeric(as.factor( CricData$Franchise)),CricData$Specialty,decreasing = FALSE),]

#tapply splitting into multiple groups or aggregating data in 3 dimensions
outputTapply<-tapply(CricData$Price, list(Country=CricData$Country, Specialty=CricData$Specialty,GFranchise=CricData$Franchise),sum)
outputTapply
outputTapply[1:10,1:5,"Mumbai"]

dt<-as.data.frame( outputTapply[1:8,1:5,"Mumbai"])
dt

(outputTapply[,,3])

dim(outputTapply)[3]

d<-lapply(1:dim(outputTapply)[3],function (x) {
  outputTapply[,,x]
})
d

d<-lapply(setNames(1:dim(outputTapply)[3],UniqueFranchise),function (x) {
  outputTapply[,,x]
})
d$Bangalore

#ftable function multidimension output to flat table.
ftable(outputTapply)

#xtabs function do cross tab sum
OutputxTab<-xtabs( CricData$Price ~ CricData$Country + CricData$Franchise + CricData$Specialty  ,CricData)
OutputxTab
OutputtApplyXTab<-tapply(CricData$Price, list(Country=CricData$Country,GFranchise=CricData$Franchise, Specialty=CricData$Specialty),sum)
ftable(OutputtApplyXTab)

ftable(OutputxTab)
#aggregate function
aggregate(CricData$Price,by=list(CricData$Franchise,CricData$Specialty), FUN=mean,na.rm=TRUE)
aggregate(CricData$Price ~ CricData$Country +CricData$Franchise+ CricData$Specialty,CricData,sum)

click to download cricdata