根据累计和和组创建新组 [英] Create new group based on cumulative sum and group

问题描述

我希望根据两个条件创建一个新组。我希望所有情况下，直到价值的累加总和达到10为止，并希望每个人都做到这一点。我设法使其分别适用于每个条件，但不能同时使用for循环和dplyr。但是，我需要同时应用这两个条件。以下是我希望数据显示的样子（我不需要RunningSum_Value列，但是为了澄清起见，我保留了它）。理想情况下，我需要dplyr解决方案，但我不挑剔。

I am looking to create a new group based on two conditions. I want all of the cases until the cumulative sum of Value reaches 10 to be grouped together and I want this done within each person. I have managed to get it to work for each of the conditions separately, but not together using for loops and dplyr. However, I need both of these conditions to be applied. Below is what I would like the data to look like (I don't need an RunningSum_Value column, but I kept it in for clarification). Ideally I would like a dplyr solution, but I m not picky. Thank you in advance!

ID     Value RunningSum_Value   Group
PersonA    1    1                 1
PersonA    3    4                 1
PersonA    10   14                1
PersonA    3    3                 2
PersonB    11   11                3
PersonB    12   12                4
PersonC    3    3                 5
PersonD    4    4                 6
PersonD    9    13                6
PersonD    5    5                 7
PersonD    11   16                7
PersonD    6    6                 8
PersonD    1    7                 8

这是我的数据：

df <- read.table(text="ID   Value
                 PersonA    1
                 PersonA    3
                 PersonA    10
                 PersonA    3
                 PersonB    11
                 PersonB    12
                 PersonC    3
                 PersonD    4
                 PersonD    9
                 PersonD    5
                 PersonD    11
                 PersonD    6
                 PersonD    1", header=TRUE,stringsAsFactors=FALSE)

推荐答案

定义函数 sum0 sum 在其参数上为 sum ，但每次达到10或更多时，其输出为0。定义函数 is_start 返回TRUE为组的起始位置，否则为FALSE。最后使用 ave 将 is_start 应用于每个ID组，然后执行 cumsum

Define function sum0 which does a sum on its argument except that each time it gets to 10 or more it outputs 0. Define function is_start that returns TRUE for the start position of a group and FALSE otherwise. Finally apply is_start to each ID group using ave and then perform a cumsum on that to get the group numbers.

sum0 <- function(x, y) { if (x + y >= 10) 0 else x + y }
is_start <- function(x) head(c(TRUE, Reduce(sum0, init=0, x, acc = TRUE)[-1] == 0), -1)
cumsum(ave(DF$Value, DF$ID, FUN = is_start))
## [1] 1 1 1 2 3 4 5 6 6 7 7 8 8

更新：修复

这篇关于根据累计和和组创建新组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！