[英]R sweep on a Sparse Matrix
I'm attempting to apply the sweep
function to a sparse matrix ( dgCMatrix
). 我正在尝试将sweep
功能应用于稀疏矩阵( dgCMatrix
)。 Unfortunately, when I do that I get a memory error. 不幸的是,当我这样做时,我遇到了内存错误。 It seems that sweep is expanding my sparse matrix to a full dense matrix. 扫描似乎将我的稀疏矩阵扩展为一个完整的密集矩阵。
If there an easy way to perform this function without if blowing up my memory? 是否有一种简便的方法来执行此功能而不会消耗我的内存?
This is what I'm trying to do. 这就是我想要做的。
sparse_matrix <- sweep(sparse_matrix, 1, vector_to_multiply, '*')
I second @user20650's recommendation to use direct multiplication of the form mat * vec
which multiplies every column of your matrix mat
with your vector vec
by implicitly recycling vec
. 我第二个@ user20650的建议是使用mat * vec
形式的直接乘法,它通过隐式回收vec
将矩阵mat
每一列与向量vec
相乘。
I understand that you're main requirement here is memory, but it's interesting to perform a microbenchmark
comparison of the sweep
and direct multiplication methods for both a dense and sparse matrix: 我了解您对内存的主要要求是,但对密集和稀疏矩阵执行sweep
和直接乘法的微microbenchmark
比较很有趣:
# Sample data
library(Matrix)
set.seed(2018)
mat <- matrix(sample(c(0, 1), 10^6, replace = T), nrow = 10^3)
mat_sparse <- Matrix(mat, sparse = T)
vec <- 1:dim(mat)[1]
library(microbenchmark)
res <- microbenchmark(
sweep_dense = sweep(mat, 1, vec, '*'),
sweep_sparse = sweep(mat_sparse, 1, vec, '*'),
mult_dense = mat * vec,
mult_sparse = mat_sparse * vec
)
res
Unit: milliseconds
expr min lq mean median uq max
sweep_dense 8.639459 10.038711 14.857274 13.064084 18.07434 32.2172
sweep_sparse 116.649865 128.111162 162.736864 135.932811 155.63415 369.3997
mult_dense 2.030882 3.193082 7.744076 4.033918 7.10471 184.9396
mult_sparse 12.998628 15.020373 20.760181 16.894000 22.95510 201.5509
library(ggplot2)
autoplot(res)
On average the operations involving a sparse matrix are actually slightly slower than the ones with a dense matrix. 平均而言,涉及稀疏矩阵的运算实际上比具有密集矩阵的运算稍慢。 Note however, how direct multiplication is faster than sweep
. 但是请注意,直接乘法比sweep
更快。
We can use memprof
to profile the memory usage of the different approaches. 我们可以使用memprof
来分析不同方法的内存使用情况。
library(profmem)
mem <- list(
sweep_dense = profmem(sweep(mat, 1, vec, '*')),
sweep_sparse = profmem(sweep(mat_sparse, 1, vec, '*')),
mult_dense = profmem(sweep(mat * vec)),
mult_sparse = profmem(sweep(mat_sparse * vec)))
lapply(mem, function(x) utils:::format.object_size(sum(x$bytes), units = "Mb"))
#$sweep_dense
#[1] "15.3 Mb"
#
#$sweep_sparse
#[1] "103.1 Mb"
#
#$mult_dense
#[1] "7.6 Mb"
#
#$mult_sparse
#[1] "13.4 Mb"
To be honest, I'm surprised that the memory imprint of the direct multiplication with a sparse matrix is not smaller than that involving a dense matrix. 老实说,我很惊讶稀疏矩阵与直接乘法的记忆烙印不小于密集矩阵。 Perhaps the sample data are too simplistic. 样本数据也许太简单了。 It might be worth exploring this with your actual data (or a representative subset thereof). 可能值得用您的实际数据(或其代表子集)进行探索。
I'm working with a big and very sparse dgTMatrix
matrix (200k rows and 10k columns) in a NLP problem. 我正在处理NLP问题中的大型且非常稀疏的dgTMatrix
矩阵(200k行和10k列)。 After hours thinking in a good solution, I created an alternative sweep
function for sparse matrices. 经过数小时的思考,找到了一个好的解决方案,我为稀疏矩阵创建了一个替代的sweep
函数。 It is very fast and memory efficient. 它非常快速且内存高效。 It took just 1 second and less than 1G of memory to multiply all matrix rows by a array of weights. 将所有矩阵行乘以权重数组仅需1秒且不到1G的内存。 For margin = 1
it works for both dgCMatrix
and dgTMatrix
. 对于margin = 1
它对dgCMatrix
和dgTMatrix
都dgTMatrix
。
Here it follows: 如下所示:
sweep_sparse <- function(x, margin, stats, fun = "*") {
f <- match.fun(fun)
if (margin == 1) {
idx <- x@i + 1
} else {
idx <- x@j + 1
}
x@x <- f(x@x, stats[idx])
return(x)
}
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.