每个组两个日期之间的平均时差

Question

我有一个包含用户及其访问日期的数据框。 我试图找出每组访问之间的平均时间差。 输出将以天或一天的一小部分为单位

require(lubridate)
so <- data.frame(visit_dates = c("12/4/2016","12/6/2016","12/7/2016","12/3/2016","12/7/2016","12/10/2016"), person = c("1","1","1","2","2","2"))


so$visit_dates <- mdy(format(as.POSIXct(strptime(so$visit_dates,"%m/%d/%Y",tz = "")),format = "%m/%d/%Y"))

输出将类似于：

person    avgTimeBetweenVisit
1                 2.5
2                 3.5

Answer 1

那个怎么样：

so %>%
  group_by(person)%>% 
  mutate(Difference = visit_dates - lag(visit_dates)) %>%
  summarize(mean_time = mean(Difference, na.rm=TRUE))

该链接帮助我解决了“差异”问题。 组内的差异操作，在 dplyr::group_by() 之后

Answer 2

试试数据表：

require(lubridate)
require(data.table)
so <- data.frame(visit_dates = c("12/4/2016","12/6/2016","12/7/2016","12/3/2016","12/7/2016","12/10/2016"), person = c("1","1","1","2","2","2"))


so$visit_dates <- mdy(format(as.POSIXct(strptime(so$visit_dates,"%m/%d/%Y",tz = "")),format = "%m/%d/%Y"))
so <- data.table(so, key = c("person", "visit_dates"))
res <- so[, .(avgTimeBetweenVisit = mean(diff(visit_dates))), by = person]
print(res)
# person avgTimeBetweenVisit
# 1:      1            1.5 days
# 2:      2            3.5 days