问题描述
如何在不将每个文件读入内存的情况下,将多个结构相同的 Arrow 文件连接成一个 Arrow 文件?我正在使用 Arrow.jl 和 Arrow 文件表示具有相同结构的数据帧,并且组合的数据帧大于可用内存。我一直在做以下事情:
function openhhmembershipchunk(basedirectory::String,node::String,chunk::Int64)
open(joinpath(basedirectory,node,"DayExtraction","HouseholdMembershipDays$(chunk).arrow")) do io
return Arrow.Table(io) |> DataFrame
end;
end
"Concatenate household membership day batches"
function combinehhmembershipdaybatch(basedirectory::String,batches)
d = openhhmembershipchunk(basedirectory::String,1)
days = copy(d)
for i = 2:batches
m = openhhmembershipchunk(basedirectory::String,i)
append!(days,m)
end
Arrow.write(joinpath(basedirectory,"HouseholdMembershipDays.arrow"),days,compress=:zstd)
#delete chunks
for i = 1:batches
rm(joinpath(basedirectory,"HouseholdMembershipDays$(i).arrow"))
end
return nothing
end #combinehhresidencydaybatch
解决方法
暂无找到可以解决该程序问题的有效方法,小编努力寻找整理中!
如果你已经找到好的解决方法,欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@)