Question

我有for for循环运行拆分尺寸的字符列（长x宽x深）然后将这些值分配给新列。现在，我遍历我的数据帧中的每个观察以产生这些变量。这是我的代码（有效），但速度很慢。

library(stringi)

for(i in 1:nrow(data3)){
  data3$Length[i] <- stri_split_fixed(str = as.character(data3$Dimensions[i]),pattern = "x", omit_empty = NA)[[1]][1]
  data3$Width[i] <- stri_split_fixed(str = as.character(data3$Dimensions[i]),pattern = "x", omit_empty = NA)[[1]][2]
  data3$Depth[i] <- stri_split_fixed(str = as.character(data3$Dimensions[i]),pattern = "x", omit_empty = NA)[[1]][3]
}

是否有可以加速此操作的plyr或apply函数？如果是这样，语法是什么？

更新：样本数据：

structure(list(Shape = structure(c(24L, 24L, 24L, 24L, 24L, 24L, 
24L, 24L, 24L, 24L), .Label = c("Asscher", "Baguette", "Briolette", 
"Bullets", "Circular Brilliant", "Cushion", "Emerald", "Flanders", 
"Half Moon", "Heart", "Hexagon", "Kite", "Lozenge", "Marquise", 
"Octagonal", "Old European", "Old Miner", "Other", "Oval", "Pear", 
"Princess", "Radiant", "Rose Cut", "Round", "Shield", "Square", 
"Tapered Baguette", "Trapezoid", "Triangular"), class = "factor"), 
    Carats = c(0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4
    ), Color = structure(c(6L, 7L, 2L, 4L, 6L, 1L, 1L, 7L, 8L, 
    2L), .Label = c("D", "E", "F", "G", "H", "I", "J", "K", "L", 
    "M", "N", "O", "P", "Q", "R", "S", "T", "U", "V", "W", "X", 
    "Y", "Z"), class = "factor"), Dimensions = c("4.58x4.61x2.97", 
    "4.66x4.69x2.95", "4.71x4.75x2.93", "4.73x4.74x2.91", "4.62x4.67x2.93", 
    "4.79x4.82x2.91", "4.6x4.62x2.92", "4.7x4.73x2.93", "4.66x4.7x2.91", 
    "4.68x4.71x2.92")), .Names = c("Shape", "Carats", "Color", 
"Dimensions"), row.names = 400:409, class = "data.frame")

Answer 1

无需循环。该函数是矢量化的，这意味着它可以立即在整个列上运行。实际上，您可以一次性获取所有三列：

data3[,c("Length","Width","Depth")] =  
          do.call(rbind, stri_split_fixed(str = as.character(data3$Dimensions), 
                                          pattern = "x", omit_empty = NA))

    Shape Carats Color     Dimensions Length Width Depth
400 Round    0.4     I 4.58x4.61x2.97   4.58  4.61  2.97
401 Round    0.4     J 4.66x4.69x2.95   4.66  4.69  2.95
402 Round    0.4     E 4.71x4.75x2.93   4.71  4.75  2.93
403 Round    0.4     G 4.73x4.74x2.91   4.73  4.74  2.91
404 Round    0.4     I 4.62x4.67x2.93   4.62  4.67  2.93
405 Round    0.4     D 4.79x4.82x2.91   4.79  4.82  2.91
406 Round    0.4     D  4.6x4.62x2.92    4.6  4.62  2.92
407 Round    0.4     J  4.7x4.73x2.93    4.7  4.73  2.93
408 Round    0.4     K  4.66x4.7x2.91   4.66   4.7  2.91
409 Round    0.4     E 4.68x4.71x2.92   4.68  4.71  2.92

Plyr /应用代替使用For循环

1 个答案: