r - 如何根据R中的多个值找到确切的变化年份

Question

我有一个包含多个主题(公司)、年份、个人姓名和性别(女性、男性)的数据框。我想获取个人姓名更改的年份(如果有更改)。此外，如果在特定年份发生变化，我想创建两个二元变量:“FemaletoMale”(和 MaletoFemale)，表示从女性到男性(男性到女性)发生的变化。

所以，如果我有一张像 这样的 table

companyid year   personalname gender 
 1         1990  Alison       Female
 1         1991  Alison       Female
 1         1992  Kate         Female
 1         1993  Kate         Female
 2         1990  George       Male
 2         1991  Kate         Female
 2         1992  Kate         Female
 3         1990  Michael      Male
 3         1991  Dwight       Male

我知道这个问题可以帮助我计算更改的数量:
How to tell if a value changed over dimensions in R df<- df %>% group_by(companyid) %>% summarise(ChangeYear = sprintf("%s to %s", min(year), max(year)), change.count = length(unique(personalname)) - 1)这给了我变化的数量。我想看到的是；

companyid  change.count  changeyear  FemaletoMale MaletoFemale
 1               1             1992         0            0          
 2               1             1991         0            1
 3               1             1991         0            0

Answer 1

这是否有效:

library(dplyr)
df %>% group_by(companyid) %>% 
  summarise(change.count = n_distinct(personalname)-1, 
         changeyear = year[personalname != lag(personalname)][2], 
         FemaletoMale = case_when(gender[1] == 'Female' & gender[n()] == 'Male' ~ 1, TRUE ~ 0),
         MaletoFemale = case_when(gender[1] == 'Male' & gender[n()] == 'Female' ~ 1, TRUE ~ 0))
`summarise()` ungrouping output (override with `.groups` argument)
# A tibble: 3 x 5
  companyid change.count changeyear FemaletoMale MaletoFemale
      <dbl>        <dbl>      <dbl>        <dbl>        <dbl>
1         1            1       1992            0            0
2         2            1       1991            0            1
3         3            1       1991            0            0