library(plyr)
nrows <- c(1e3, 5e3, 1e4 ,5e4, 1e5 ,2e5)
res<-list()

for(i in nrows){
  print(i)
  dum <- data.frame(iddd = 1:i, x = runif(i,1,1000), y=runif(i,1,1000))
  res[[length(res)+1]]<-(system.time(x<- dlply(dum, .variables=.(iddd),.parallel=FALSE)))
}
res <- do.call("rbind",res)
plot(nrows, res[,"elapsed"])
plot(nrows*log(nrows), res[,"elapsed"])

datasplit <- function(x){
  res <- vector(mode="list", length= nrow(x))
  for(i in 1:nrow(x)){
    res <-  x[i,] 
  }
  return(res)
}

res <- datasplit(dum)

for(i in nrows){
  print(i)
  dum <- data.frame(iddd = 1:i, x = runif(i,1,1000), y=runif(i,1,1000))
  res[[length(res)+1]]<-(system.time(x<- dlply(dum, .variables=.(iddd),.parallel=FALSE)))
}
res <- do.call("rbind",res)
plot(nrows, res[,"elapsed"])
plot(nrows*log(nrows), res[,"elapsed"])


nrows <- c(1e3,5e3, 1e4 ,5e4, 1e5 ,2e5)
res<-list()

for(i in nrows){
  print(i)
  dum <- data.frame(x = runif(i,1,1000), y=runif(i,1,1000))
  res[[length(res)+1]]<-(system.time(x<- split(dum, 1:nrow(dum))))
}
res2 <- do.call("rbind",res)
plot(nrows^2, res2[,"elapsed"])

nrows <- c(1e3,5e3, 1e4 ,5e4, 1e5 ,2e5, 1e6, 2e6)
res <- list()
for(i in nrows){
  print(i)
  dum <- data.frame(iddd = 1:i, x = runif(i,1,1000), y=runif(i,1,1000))
  print(dim(dum))
  idx <- sample( i, 10000,replace = TRUE )
  print(length(idx))
  res[[length(res)+1]]<-(system.time(x<- dum[idx,]))
}
res3 <- do.call("rbind",res)
plot(nrows, res3[,"elapsed"])


protViz/bibliospec documentation built on May 26, 2019, 9:37 a.m.