按行选择第一个不缺少的元素 [英] Pick out first non missing element by row

问题描述

假设我想要执行以下操作：

Suppose I want to do the following:

1）应用一些函数，它们用空格逐行替换不需要的字符;
2）然后它选择（从原始数据中）逐行第一个非空元素（根据上面的转换）。

1) Apply some function, which replaces not needed characters by empty spaces row by row; 2) Then it picks out (out of the original data) row by row first non empty element (according to the above transformation).

这里是我现在的：

library(data.table)
data<-data.table(x=c("25&&&35&&1","&&&&","&&&&"),
                 y=c("&&&&&","1","&&&&2"),
                 z=c("&&&&&","","1"))

function_select<-function(x){
  x[gsub("&","",x)!=""][1]
}

data[,function_select(unlist(.SD)),.SDcols=c("x","y"),by=1:nrow(data)]

   nrow         V1
1:    1 25&&&35&&1
2:    2          1
3:    3      &&&&2

我有大约7000万行，问题：
1）使用data.table有更快的解决方案;
2）我可以使用parLapply的想法加快计算速度;

I have roughly 70 millions of rows, I have two questions: 1) Is there a faster solution in using data.table; 2) Can I speed up my calculations using ideas from parLapply;

推荐答案

鼠疫。这里我可能会：

data[, lapply(.SD, sub, pattern = "^&*$", replacement = ""), .SDcols = x:y][
     , as.matrix(.SD)[.SD != ""]]
#[1] "25&&&35&&1" "1"          "&&&&2"

这篇关于按行选择第一个不缺少的元素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！