使用R计算中位数而不复制元素

Use R to calculate median without replicating elements

我的频率分布很大。我想计算中位数和四分位数,但 R 抱怨。以下是适用于小数字的方法:

> TABLE <- data.frame(DATA = c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19), F = c(48,0,192,1152,5664,23040,77952,214272,423984,558720,267840,0,0,0,0,0,0,0,0))
> summary(rep(TABLE$DAT,TABLE$F))
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  1.000   9.000  10.000   9.397  10.000  11.000

这是我得到的大数字:

> TABLE <- data.frame(DATA = c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19), F = c(240,0,1200,9600,69600,470400,2992800,17859840,98312880,489292800,2164619760,8325820800,26865302400,68711068800,128967422400,153763315200,96770419200,26824089600,2395008000))
> summary(rep(TABLE$DAT,TABLE$F))
Error in rep(TABLE$DAT, TABLE$F) : invalid 'times' argument
In addition: Warning message:
In summary(rep(TABLE$DAT, TABLE$F)) :
  NAs introduced by coercion to integer range

这个错误并不让我感到意外,因为使用 "rep" 我想创建一个巨大的向量。但我不知道,如何避免这种情况并计算中位数和四分位数。

与其试图复制那个怪物来使用 summary(),不如得到 "weighted quantiles"。 This post has a formula。 但与大多数事情一样,一旦你知道了正确的条款,你就可以找到一个包裹 那已经完成了工作!

#install.packages("Hmisc")

TABLE <- data.frame(DATA = c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19), F = c(240,0,1200,9600,69600,470400,2992800,17859840,98312880,489292800,2164619760,8325820800,26865302400,68711068800,128967422400,153763315200,96770419200,26824089600,2395008000))


Hmisc::wtd.quantile(TABLE$DATA, probs = c(0.25, 0.5, 0.75), weight = TABLE$F)
#> 25% 50% 75% 
#>  15  16  16

reprex package (v0.2.0) 创建于 2018-04-06。