1

我已经把一个完美的工作map变成了一个pmap

squashed_groups = pmap(group ->
                       reduce(squash,
                              DataFrame(eltypes(o), names(o), 0),
                              eachrow(group)
                             ),
                       groups)

除了它似乎没有参与任何并行性这一事实 - 但我承认这些组大多很小 - 我收到以下错误

ERROR: LoadError: type DataFrame has no field x
Stacktrace:
 [1] copy!(::Array{DataFrames.AbstractDataFrame,1}, ::Base.Generator{Array{DataFrames.AbstractDataFrame,1},DataFrames.##34#35{Base.##569#571}}) at ./abstractarray.jl:572
 [2] _collect(::Type{DataFrames.AbstractDataFrame}, ::Base.Generator{Array{DataFrames.AbstractDataFrame,1},DataFrames.##34#35{Base.##569#571}}, ::Base.HasShape) at ./array.jl:363
 [3] map(::Base.##569#571, ::DataFrames.GroupApplied) at /Users/morpheu5/.julia/v0.6/DataFrames/src/groupeddataframe/grouping.jl:184
 [4] maptwice(::Function, ::Channel{Any}, ::Array{Any,1}, ::DataFrames.GroupedDataFrame, ::Vararg{DataFrames.GroupedDataFrame,N} where N) at ./asyncmap.jl:188
 [5] wrap_n_exec_twice(::Channel{Any}, ::Array{Any,1}, ::Base.Distributed.##204#207{WorkerPool}, ::Function, ::DataFrames.GroupedDataFrame, ::Vararg{DataFrames.GroupedDataFrame,N} where N) at ./asyncmap.jl:154
 [6] #async_usemap#553(::Function, ::Void, ::Function, ::##9#10, ::DataFrames.GroupedDataFrame, ::Vararg{DataFrames.GroupedDataFrame,N} where N) at ./asyncmap.jl:103
 [7] (::Base.#kw##async_usemap)(::Array{Any,1}, ::Base.#async_usemap, ::Function, ::DataFrames.GroupedDataFrame, ::Vararg{DataFrames.GroupedDataFrame,N} where N) at ./<missing>:0
 [8] (::Base.#kw##asyncmap)(::Array{Any,1}, ::Base.#asyncmap, ::Function, ::DataFrames.GroupedDataFrame) at ./<missing>:0
 [9] #pmap#203(::Bool, ::Int64, ::Void, ::Array{Any,1}, ::Void, ::Function, ::WorkerPool, ::Function, ::DataFrames.GroupedDataFrame) at ./distributed/pmap.jl:126
 [10] pmap(::WorkerPool, ::Function, ::DataFrames.GroupedDataFrame) at ./distributed/pmap.jl:101
 [11] #pmap#213(::Array{Any,1}, ::Function, ::Function, ::DataFrames.GroupedDataFrame) at ./distributed/pmap.jl:156
 [12] pmap(::Function, ::DataFrames.GroupedDataFrame) at ./distributed/pmap.jl:156
 [13] include_from_node1(::String) at ./loading.jl:569
 [14] include(::String) at ./sysimg.jl:14
 [15] process_options(::Base.JLOptions) at ./client.jl:305
 [16] _start() at ./client.jl:371

我想不通那是什么意思。我什至检查了我可以在某个地方使用的任何x地方并更改了这些地方，但错误消息没有改变。内部 reduce 有点复杂，但可以使用非并行来工作map——例如，它会产生预期的结果。

这是我阅读了所有文档但没有理解它们的情况，还是这是 Julia 中的某个错误？我在 x86_64-apple-darwin13.4.0 上运行 0.6.0。

4

0 回答 0