Introdução

Pacotes utilizados neste documento:

library(Rcpp)
library(tidyr)
library(dplyr, warn.conflicts =  FALSE)
library(ggplot2)
library(microbenchmark)

Funções

C++ usando o pacote “Rcpp”.

#include <Rcpp.h>
using namespace Rcpp;

// [[Rcpp::export]]
NumericVector rowSumsC(NumericMatrix x) {
  int nrow = x.nrow(), ncol = x.ncol();
  NumericVector out(nrow);

  for (int i = 0; i < nrow; i++) {
    double total = 0;
    for (int j = 0; j < ncol; j++) {
      total += x(i, j);
    }
    out[i] = total;
  }
  return out;
}
cppSum <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  Time <- microbenchmark({
    colsums <- rowSumsC(big.matrix)
  }, times = N)$time
  return(Time)  
}

Usando a função otimizada “colSums” do R:

ColSum <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  Time <- microbenchmark({
    colsums <- colSums(big.matrix)
  }, times = N)$time
  return(Time)  
}

Usando função “apply” nativa do R:

Apply <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  Time <- microbenchmark({
    colsums <- apply(big.matrix, 2, sum)
  }, times = N)$time
  return(Time)  
}

Usando multiplicação matricial:

matrixSum <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  Time <- microbenchmark({
    ones <- rep(1, dim(big.matrix)[1])
    colsums <- ones %*% big.matrix
  }, times = N)$time
  return(Time)  
}

Função com Laço simple:

SingleLoop <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  colsums <- rep(NA, dim(big.matrix)[2])
  Time <- microbenchmark({
    for (i in 1:dim(big.matrix)[2]){
      colsums[i] <- sum(big.matrix[,i])
    }
  }, times = N)$time
  return(Time)
}

Função com laço duplo:

DoubleLoop <- function(big.matrix, N) {
  stopifnot(exists("big.matrix"))
  stopifnot(exists("N"))
  Time <- microbenchmark({
  colsums <- rep(NA, dim(big.matrix)[2])    
      for (i in 1:dim(big.matrix)[2]){
        s <- 0
        for (j in 1:dim(big.matrix)[1]){
          s <- s + big.matrix[j, i]
        }
        colsums[i] <- s
      }
    }, times = N)$time
  return(Time)
} 

Primeiro Experimento

Trials = 100 # Number of experiments
Experiments <- as.data.frame(matrix(NA, nrow= Trials, ncol=5))
colnames(Experiments)<-c("CppSum", "SingleLoop", "Apply", "ColSum", "MatrixSum")

for (n in 1:Trials) {
  big.matrix <- matrix(runif(1e+6), nrow = 1000)
  Experiments[n,] = c(cppSum(big.matrix, N = 1),
                      SingleLoop(big.matrix, N = 1), 
                      Apply(big.matrix, N = 1), 
                      ColSum(big.matrix, N = 1), 
                      matrixSum(big.matrix, N = 1))
}

p  <-  Experiments  %>% gather("Método", "Tempo")  %>% ggplot(aes(x =  reorder(Método, Tempo, FUN=median), y = Tempo)) 
p + geom_boxplot() + labs(x = "Método", y = "Tempo em nanosegundos")

Segundo Experimento

Trials = 10 # Number of experiments
Sizes = c(1000, 2000, 4000, 6000)

ExperimentsSizes <- as.data.frame(matrix(NA, nrow= Trials*length(Sizes), ncol=6))
colnames(ExperimentsSizes)<-c("CppSum", "SingleLoop", "Apply", "ColSum", "MatrixSum", "Size")
i <- 1
for (s in Sizes){
  for (n in 1:Trials) {
  big.matrix <- matrix(runif(s*s), nrow = s)
  ExperimentsSizes[i,] = c(cppSum(big.matrix, N = 1),
                      SingleLoop(big.matrix, N = 1), 
                      Apply(big.matrix, N = 1), 
                      ColSum(big.matrix, N = 1), 
                      matrixSum(big.matrix, N = 1),
                      s)
  i <- i + 1
}  
}

p<- ExperimentsSizes  %>% gather("Método", "Tempo", -Size)  %>% ggplot(aes(x = factor(Size), y = Tempo, fill = reorder(factor(Método), Tempo, FUN=median)))
p + geom_boxplot() + labs(fill = "Método", x = "Tamanho da Matriz", y = "Tempo em nanosegundos")