获得一个人居住的第一个位置

Posted 2021-04-25

我想根据年份栏创建一个人居住的第一栏。我有以下格式的数据

year<- c(2008, 2009, 2010, 2009, 2010, 2011)
person<- c('John', 'John', 'John', 'Brian', 'Brian','Vickey')
location<- c('London','Paris', 'Newyork','Paris','Paris','Miami')
df<- data.frame(year, person, location)

我想创建一个名为first place的列，其值为0,1。如果它是第一个城市，则为1，否则为0。

有什么建议？

答案

与dplyr的解决方案：

df %>% 
    group_by(person) %>% 
    mutate(FirstPlace = +(location[which.min(year)] == location))

# A tibble: 6 x 4
# Groups:   person [3]
#   year person location FirstPlace
#  <dbl> <fctr>   <fctr>      <int>
#1  2008   John   London          1
#2  2009   John    Paris          0
#3  2010   John  Newyork          0
#4  2009  Brian    Paris          1
#5  2010  Brian    Paris          1
#6  2011 Vickey    Miami          1

• location[which.min(year)]找出第一个位置，然后将第一个位置与位置列进行比较，并将布尔结果转换为整数。

如果只看第一年：

df %>% 
    group_by(person) %>% 
    mutate(FirstPlace = +(min(year) == year))

# A tibble: 6 x 4
# Groups:   person [3]
#   year person location FirstPlace
#  <dbl> <fctr>   <fctr>      <int>
#1  2008   John   London          1
#2  2009   John    Paris          0
#3  2010   John  Newyork          0
#4  2009  Brian    Paris          1
#5  2010  Brian    Paris          0
#6  2011 Vickey    Miami          1

另一答案

与data.tableyou可以做：

library("data.table")
year<- c(2008, 2009, 2010, 2009, 2010, 2011)
person<- c('John', 'John', 'John', 'Brian', 'Brian','Vickey')
location<- c('London','Paris', 'Newyork','Paris','Paris','Miami')
df<- data.frame(year, person, location)
setDT(df)[, firstPlace:=as.integer(min(year)==year), person]
# > setDT(df)[, firstPlace:=as.integer(min(year)==year), person]
# > df
#    year person location firstPlace
# 1: 2008   John   London          1
# 2: 2009   John    Paris          0
# 3: 2010   John  Newyork          0
# 4: 2009  Brian    Paris          1
# 5: 2010  Brian    Paris          0
# 6: 2011 Vickey    Miami          1

或者（如@Frank提到的）您的数据是按人和年分类的

setDT(df)[, firstPlace:=+!duplicated(person)]

或（此变体）

setDT(df)[, firstPlace:=+(rowidv(person)==1)]

另一答案

first_city<-df%>%
group_by(person)%>%
  arrange(year)%>%
  slice(1)

另一答案

library(dplyr)
first_city <- df %>%
  group_by(person) %>%
  top_n(1, year)

或者作为数据中的额外列：

df %>% 
  group_by(person) %>% 
  arrange(year) %>% 
  mutate(first_city = head(location, 1))

指示第一个城市有1和0 otherwuse（仅限第一年）

df %>% 
  group_by(person) %>% 
  arrange(year) %>% 
  mutate(first_city = as.integer(head(location, 1) == location & year == min(year))) 

# A tibble: 6 x 4
# Groups:   person [3]
#    year person location first_city
#   <dbl> <fct>  <fct>         <int>
# 1  2008 John   London            1
# 2  2009 John   Paris             0
# 3  2009 Brian  Paris             1
# 4  2010 John   Newyork           0
# 5  2010 Brian  Paris             0
# 6  2011 Vickey Miami             1