Vit*_*ijs 6 parallel-processing r pmap julia mclapply
我正在尝试将我的一些 R 代码移植到 Julia;基本上我已经在 Julia 中重写了以下 R 代码:
library(parallel)
eps_1<-rnorm(1000000)
eps_2<-rnorm(1000000)
large_matrix<-ifelse(cbind(eps_1,eps_2)>0,1,0)
matrix_to_compare = expand.grid(c(0,1),c(0,1))
indices<-seq(1,1000000,4)
large_matrix<-lapply(indices,function(i)(large_matrix[i:(i+3),]))
function_compare<-function(x){
which((rowSums(x==matrix_to_compare)==2) %in% TRUE)
}
> system.time(lapply(large_matrix,function_compare))
user system elapsed
38.812 0.024 38.828
> system.time(mclapply(large_matrix,function_compare,mc.cores=11))
user system elapsed
63.128 1.648 6.108
Run Code Online (Sandbox Code Playgroud)
正如人们所注意到的,当从 1 个内核变为 11 个内核时,我获得了显着的加速。现在我正尝试在 Julia 中做同样的事情:
#Define cluster:
addprocs(11);
using Distributions;
@everywhere using Iterators;
d = Normal();
eps_1 = rand(d,1000000);
eps_2 = rand(d,1000000);
#Create a large matrix:
large_matrix = hcat(eps_1,eps_2).>=0;
indices = collect(1:4:1000000)
#Split large matrix:
large_matrix = [large_matrix[i:(i+3),:] for i in indices];
#Define the function to apply:
@everywhere function function_split(x)
matrix_to_compare = transpose(reinterpret(Int,collect(product([0,1],[0,1])),(2,4)));
matrix_to_compare = matrix_to_compare.>0;
find(sum(x.==matrix_to_compare,2).==2)
end
@time map(function_split,large_matrix )
@time pmap(function_split,large_matrix )
5.167820 seconds (22.00 M allocations: 2.899 GB, 12.83% gc time)
18.569198 seconds (40.34 M allocations: 2.082 GB, 5.71% gc time)
Run Code Online (Sandbox Code Playgroud)
正如人们所注意到的,我没有使用 pmap 加快速度。也许有人可以提出替代方案。
我认为这里的一些问题是,@parallel并不@pmap总是能很好地处理与工作人员之间的数据移动。因此,它们往往在您所执行的操作根本不需要太多数据移动的情况下工作得最好。我还怀疑可能可以采取一些措施来提高他们的表现,但我不确定细节。
对于确实需要移动更多数据的情况,最好坚持使用直接调用工作人员函数的选项,然后这些函数访问这些工作人员内存空间内的对象。我在下面给出了一个例子,它可以使用多个工作人员来加速您的功能。它使用也许是最简单的选项,即@everywhere,但是@spawn等remotecall()也值得考虑,具体取决于您的情况。
addprocs(11);
using Distributions;
@everywhere using Iterators;
d = Normal();
eps_1 = rand(d,1000000);
eps_2 = rand(d,1000000);
#Create a large matrix:
large_matrix = hcat(eps_1,eps_2).>=0;
indices = collect(1:4:1000000);
#Split large matrix:
large_matrix = [large_matrix[i:(i+3),:] for i in indices];
large_matrix = convert(Array{BitArray}, large_matrix);
function sendto(p::Int; args...)
for (nm, val) in args
@spawnat(p, eval(Main, Expr(:(=), nm, val)))
end
end
getfrom(p::Int, nm::Symbol; mod=Main) = fetch(@spawnat(p, getfield(mod, nm)))
@everywhere function function_split(x::BitArray)
matrix_to_compare = transpose(reinterpret(Int,collect(product([0,1],[0,1])),(2,4)));
matrix_to_compare = matrix_to_compare.>0;
find(sum(x.==matrix_to_compare,2).==2)
end
function distribute_data(X::Array, WorkerName::Symbol)
size_per_worker = floor(Int,size(X,1) / nworkers())
StartIdx = 1
EndIdx = size_per_worker
for (idx, pid) in enumerate(workers())
if idx == nworkers()
EndIdx = size(X,1)
end
@spawnat(pid, eval(Main, Expr(:(=), WorkerName, X[StartIdx:EndIdx])))
StartIdx = EndIdx + 1
EndIdx = EndIdx + size_per_worker - 1
end
end
distribute_data(large_matrix, :large_matrix)
function parallel_split()
@everywhere begin
if myid() != 1
result = map(function_split,large_matrix );
end
end
results = cell(nworkers())
for (idx, pid) in enumerate(workers())
results[idx] = getfrom(pid, :result)
end
vcat(results...)
end
## results given after running once to compile
@time a = map(function_split,large_matrix); ## 6.499737 seconds (22.00 M allocations: 2.899 GB, 13.99% gc time)
@time b = parallel_split(); ## 1.097586 seconds (1.50 M allocations: 64.508 MB, 3.28% gc time)
julia> a == b
true
Run Code Online (Sandbox Code Playgroud)
注意:即使这样,多个进程的加速也不是完美的。但是,这是可以预料的,因为您的函数仍然会返回适量的数据,并且必须移动这些数据,这需要时间。
PS 请参阅这篇文章(Julia:如何将数据复制到 Julia 中的另一个处理器)或此包(https://github.com/ChrisRackauckas/ParallelDataTransfer.jl),了解有关我在这里使用的sendto和getfrom函数的更多信息。