Question

这是在两个级别使用R中的聚合函数。对于以下数据框，

df <- data.frame( Event  = c('A1','A1','A1','A1','A1'),
                    Time = c(10,11,21,17,12),
                    Type = c('New','New','Repeat','Repeat','Repeat'))

在某个级别，以下是完美的工作

aggregate(df$Time,
          by=list(df$Event), 
          FUN = function(x) c(mintime = min(x),endtime = max(x)))

我正在寻找带有条件检查的每个Type的计数。

## CODE NOT WORKING
aggregate(df,by=list(df$Event), 
          FUN = function(x) c(mintime = min(df$time),endtime = max(df$time), New=length(df$time=='New'),Repeat=length(df$time=='Repeat')))

所以我会得到

Group.1 x.mintime x.maxtime x.New x.Repeat
A1      10        21        2     3

Answer 1

使用server.contextPath=/refms：

localhost:8080/refms/api/account/authenticate

使用localhost:8080/refms-xxxx/api/account/authenticate：

data.table

<强>结果：

library(data.table)
setDT(df)[, .(mintime = min(Time), 
              maxtime = max(Time), 
              New = sum(Type=="New"), 
              Repeat = sum(Type=="Repeat")),
          by = Event]

Answer 2

使用来自dplyr的 group_by（）和 summary（），这非常容易。请参阅以下内容：

library(dplyr)

df <- data.frame( Event  = c('A1','A1','A1','A1','A1'),
                  Time = c(10,11,21,17,12),
                  Type = c('New','New','Repeat','Repeat','Repeat'))

df %>% 
  group_by(Event) %>% 
  summarise(mintime  = min(Time),
            endtime  = max(Time),
            x.New    = sum(Type == "New"),
            x.Repeat = sum(Type == "Repeat"))

# A tibble: 1 x 5
   Event mintime endtime x.New x.Repeat
  <fctr>   <dbl>   <dbl> <int>    <int>
1     A1      10      21     2        3

Aggregate（） - 多个级别

2 个答案: