哪个是并行化余弦距离的最佳方法? [英] Which is the best way to parallelize cosine distance?

问题描述

当我尝试使用大型数据集(约 600,000 行)计算余弦距离时，我的 R 会话在超时后崩溃

My R session crashes after the timeout is exceeded when I try to compute the cosine distance with a large dataset (~600,000 lines)

对于小数据集，我的代码有效，这是一个示例:

For small datasets my code works and this is an example:

library(lsa)
relevant.data <- as.matrix(mtcars)
cosine(t(relevant.data))

我读过这个网站上的一些帖子来并行化余弦函数，但没有运气.

I've read some posts on this website to parallelize cosine function but no luck.

是否存在非常有效的方法?

Does a very efficient method exist?

你推荐 rccp 喜欢这篇文章吗?在R中使用clusterapply的平行余弦距离

Do you suggest rccp like this post? Parallel cosine distance using clusterapply in R

如果计算相关矩阵之类的东西效率低下.你有什么建议?

If computing something like a correlation matrix is inefficient. What do you suggest?

推荐答案

在 Rcpp 中进行编码可能会让您感到满意，您不需要额外的并行化麻烦.下面的示例(但我不知道它会如何在您的系统上执行/遇到实际大小的问题:长度为 1e8 的向量(相当于 10,000 x 10,000 矩阵)占用 763Mb，因此即使存储问题 60 的结果^2 倍(如果我计算正确，则为 2.75Tb)可能很困难......).

Coding it in Rcpp might buy you enough that you don't need the extra hassle of parallelizing. Example below (but I don't know how it will do on your system/with a real-sized problem: a vector of length 1e8 (equivalent to a 10,000 by 10,000 matrix) takes 763Mb, so even storing the results for a problem 60^2 times larger (=2.75Tb if I've calculated correctly) might be difficult ...).

x <- as.matrix(mtcars)
library(lsa)

来自 lsa 的函数:

cosine(as.matrix(mtcars))

稍微精简的 R 代码:

Slightly stripped-down R code:

cosR <- function(x) {
      co <- array(0, c(ncol(x), ncol(x)))
      ## f <- colnames(x)
      ## dimnames(co) <- list(f, f)
      for (i in 2:ncol(x)) {
        for (j in 1:(i - 1)) {
            co[i,j] <- crossprod(x[,i], x[,j])/
                sqrt(crossprod(x[,i]) * crossprod(x[,j]))
        }
    }
    co <- co + t(co)
    diag(co) <- 1
    return(as.matrix(co))
}

Rcpp 版本，从此处稍作修改:

Rcpp version, slightly modified from here:

library(Rcpp)
library(RcppArmadillo)
cppFunction(depends='RcppArmadillo',
            code="NumericMatrix cosCpp(NumericMatrix Xr) {
            int n = Xr.nrow(), k = Xr.ncol();
            arma::mat X(Xr.begin(), n, k, false); // reuses memory and avoids extra copy
            arma::mat Y = arma::trans(X) * X; // matrix product
            arma::mat res = Y / (arma::sqrt(arma::diagvec(Y)) * arma::trans(arma::sqrt(arma::diagvec(Y))));
            return Rcpp::wrap(res);
           }")

测试相等性:

identical(cosR(x),unname(cosine(x)))
all.equal(cosCpp(x),cosR(x))

library(microbenchmark)
microbenchmark(cosine(x),cosR(x),cosCpp(x))
## Unit: nanoseconds
##       expr    min      lq       mean  median      uq      max neval cld
##  cosine(x) 460046 1181837 2069604.51 1530719 2528021  8757989   100   b
##    cosR(x) 542414 1096448 1915011.12 1331277 2321596 11740233   100   b
##  cosCpp(x)      7   12472   35827.76   17999   30556   644551   100  a 

Rcpp 版本大约快了 1331277/17999 = 74 倍，并且可能 (?) 也能让您解决内存问题.

The Rcpp version is about 1331277/17999 = 74 times faster, and might (?) get you around memory issues as well.

这篇关于哪个是并行化余弦距离的最佳方法?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！