Plot the run time of three functions

问题

I have three functions doing the same task but in a different way:

Function 1:

f1 <- function(n) {
  sum = 0
  for(i in 1:n) {
    sum <- sum + a(i)
  }
  return(sum)
}

Function 2:

f4 <- function(n) {
  sum = 0
  for(i in 1:n) {
    sum <- sum + b(i)
  }
  return(sum)
}

Function 3:

f3 <- function(n) {
  #Done using lgamma
  sum = 0
  for(i in 1:n) {
    sum = sum + c(i)
  }
  return(sum)
}

I need to plot the runtime of these 3 functions over a range of values 1 - n using user time from system.time. How can this be done in R?

回答1:

You could use microbenchmark to time your functions, and ggplot2 to plot. Here is an example:

library(microbenchmark)
library(reshape2)
library(ggplot2)

max_n = 15
results = data.frame(sum_lgamma=numeric(max_n),
                     sum_log_gamma_recursive = numeric(max_n),
                     sum_log_gamma_loop = numeric(max_n),
                     id = seq(max_n))
for(i in 1:max_n)
{
  results$sum_lgamma[i] = median(microbenchmark::microbenchmark(sum_lgamma(i))$time)
  results$sum_log_gamma_loop[i] = median(microbenchmark::microbenchmark(sum_log_gamma_loop(i))$time)
  results$sum_log_gamma_recursive[i] = median(microbenchmark::microbenchmark(sum_log_gamma_recursive(i))$time)
}
results = melt(results, id.vars=c("id"))

ggplot(results, aes(x=id,y=value,color=variable)) + geom_line()

来源：https://stackoverflow.com/questions/48474231/plot-the-run-time-of-three-functions