计算几个列案例的出现次数

Question

我有一个数据框：

ID   Date            col1  col2  
1    1606807860      LOY    A
2    1606807860      LOY    B
2    1606807860      LOY    B
3    1606807860      LOY    B
1    1606807860      LOY    A

我想根据 ID、Date、col1 和 col2 计算唯一值的出现次数。所以，想要的结果是：

ID     Date              event    count
1    1606807860          loy-a      2
2    1606807860          loy-b      2
3    1606807860          loy-b      1

我怎么能这样做？另外如何将时间戳格式转换为正常格式，而不是 1606807860？如何更改日期类型？把它变成年月日？

这适用于只有 col1 和 col2 的情况：

%>%
  mutate(across(c(col1, col2), tolower)) %>%
  count(col1, col2) %>%
  unite(event, col1, col2, sep='-')

Answer 1

在这种情况下，我们不是一一指定多个列，而是在group_by 中使用across，然后从names 和summarise 中指定一个列范围

library(dplyr)
library(stringr)
df1 %>%
   group_by(across(names(.)[1:4])) %>%
   summarise(count = n(), .groups = 'drop') %>%
   mutate(event = tolower(str_c(col1, col2, sep="-"))) %>%
   select(-col1, -col2)

-输出

# A tibble: 3 x 4
#     ID       Date count event
#  <int>      <int> <int> <chr>
#1     1 1606807860     2 loy-a
#2     2 1606807860     2 loy-b
#3     3 1606807860     1 loy-b

---

或者我们可以创建group_by 为

df1 %>% 
   group_by(across(ID:Date), event = tolower(str_c(col1, col2, sep='-'))) %>%
   summarise(count = n(), .groups = 'drop') %>%
   mutate(Date = as.Date(as.POSIXct(Date, origin = '1970-01-01')))

-输出

# A tibble: 3 x 4
#     ID Date       event count
#  <int> <date>     <chr> <int>
#1     1 2020-12-01 loy-a     2
#2     2 2020-12-01 loy-b     2
#3     3 2020-12-01 loy-b     1

---

或者count和unite

library(tidyr)
df1 %>%
   mutate(across(c(col1, col2), tolower)) %>%
   unite(event, col1, col2, sep='-') %>%
   count(ID, Date, event)

-输出

#   ID       Date event n
#1  1 1606807860 loy-a 2
#2  2 1606807860 loy-b 2
#3  3 1606807860 loy-b 1

数据

df1 <- structure(list(ID = c(1L, 2L, 2L, 3L, 1L), Date = c(1606807860L, 
1606807860L, 1606807860L, 1606807860L, 1606807860L), col1 = c("LOY", 
"LOY", "LOY", "LOY", "LOY"), col2 = c("A", "B", "B", "B", "A"
)), class = "data.frame", row.names = c(NA, -5L))

Answer 2

试试这个：

library(dplyr)
#Code
new <- df %>% group_by(ID,Date,event=tolower(paste0(col1,'-',col2))) %>%
  summarise(N=n()) %>% mutate(Date=as.Date(as.POSIXct(Date,origin = "1970-01-01")))

输出：

# A tibble: 3 x 4
# Groups:   ID, Date [3]
     ID Date       event     N
  <int> <date>     <chr> <int>
1     1 2020-12-01 loy-a     2
2     2 2020-12-01 loy-b     2
3     3 2020-12-01 loy-b     1

Answer 3

基本 R 选项

aggregate(
  n ~ .,
  transform(
    df,
    event = tolower(paste(col1, col2, sep = "-")),
    Date = as.Date(as.POSIXct(Date, origin = "1970-01-01")),
    n = 1,
    col1 = NULL,
    col2 = NULL
  ),
  sum
)

给了

  ID       Date event n
1  1 2020-12-01 loy-a 2
2  2 2020-12-01 loy-b 2
3  3 2020-12-01 loy-b 1

---

data.table 选项

setDT(df)
df[, Date := as.Date(as.POSIXct(Date, origin = "1970-01-01"))][, .(event = tolower(paste(col1, col2, sep = "-")), n = .N), by = names(df)][, c("col1", "col2") := NULL][]

给了

   ID       Date event n
1:  1 2020-12-01 loy-a 2
2:  2 2020-12-01 loy-b 2
3:  3 2020-12-01 loy-b 1

数据

> dput(df)
structure(list(ID = c(1L, 2L, 2L, 3L, 1L), Date = c(1606807860L,
1606807860L, 1606807860L, 1606807860L, 1606807860L), col1 = c("LOY",
"LOY", "LOY", "LOY", "LOY"), col2 = c("A", "B", "B", "B", "A"
)), class = "data.frame", row.names = c(NA, -5L))