主页 / 问题 / 849473
Accepted
Alexshev92
Asked:2020-07-03 12:58:31 +0000 UTC2020-07-03 12:58:31 +0000 UTC

计算 data.frame 中相同行的数量

  • 772

有一个三列和一些行数的data.frame

vendor <- c("BMW", "BMW", "BMW", "Audi", "Audi", "Mersedes", "BMW", "Audi")
model <- c("X1", "X5", "X1", "A6", "A6", "C-350", "X1", "A6")
year <- c(2009, 2011, 2010, 2015, 2015, 2011, 2010, 2016)

cars <- data.frame(vendor, model, year)

    vendor model year
1      BMW    X1 2009
2      BMW    X5 2011
3      BMW    X1 2010
4     Audi    A6 2015
5     Audi    A6 2015
6 Mersedes C-350 2011
7      BMW    X1 2010
8     Audi    A6 2016

我如何计算相同行的数量，理想情况下得到一个带有第四列的新数据框，它表示重复的次数。就像是：

    vendor model year count
1      BMW    X1 2009 1
2      BMW    X5 2011 1
3      BMW    X1 2010 2
4     Audi    A6 2015 2
5 Mersedes C-350 2011 1
6     Audi    A6 2016 1

尝试使用该功能

summary(cars)

但它输出每列的统计信息。

r

2 个回答

  1. Best Answer

    unique(cars) vendor model year 1 BMW X1 2009 2 BMW X5 2011 3 BMW X1 2010 4 Audi A6 2015 6 Mersedes C-350 2011 8 Audi A6 2016

    或计算行数

    library(data.table) cars <- data.table(cars) cars[, .N, by = names(cars)] vendor model year N 1: BMW X1 2009 1 2: BMW X5 2011 1 3: BMW X1 2010 2 4: Audi A6 2015 2 5: Mersedes C-350 2011 1 6: Audi A6 2016 1

    另见http://www.sthda.com/english/wiki/identifying-and-removing-duplicate-data-in-r

    • 3

  2. 使用包的选项dplyr

    library(dplyr)

cars.1 <- group_by(cars, vendor, model, year)
cars.2 <- summarise(cars.1, N = n())

# A tibble: 6 x 4
# Groups:   vendor, model [?]
  vendor   model  year     N
  <fct>    <fct> <dbl> <int>
1 Audi     A6     2015     2
2 Audi     A6     2016     1
3 BMW      X1     2009     1
4 BMW      X1     2010     2
5 BMW      X5     2011     1
6 Mersedes C-350  2011     1

group_by - группируем заданный набор данных по заданным столбцам
summarise - вычисляем итоги для каждой группы по строкам
n() - количество наблюдений в группе

# вариант записи без промежуточных итогов используюя оператов %>%
cars.3 <- cars %>% group_by(vendor, model, year) %>% summarise(N = n())

    dplyr俄语包介绍

    • 1

相关问题