我怎样才能更广泛地旋转并转换我的数据框？

Question

我有一个这样的数据框：

tibble(
  School = c(1, 1, 2, 3, 3, 4),
  City = c("A","A", "B", "C", "C", "B"),
  Grade = c("7th", "7th", "7th", "6th", "8th", "8th"),
  Number_Students = c(20, 23, 25, 21, 28, 34),
  Type_school = c("public", "public", "private", "public", "public", "private")
  )

ID	School	City	Grade	Number_Students	Type_school
1	1	A	7th	20	public
2	1	A	7th	23	public
3	2	B	7th	25	private
4	3	C	6th	21	public
5	3	C	8th	28	public
6	4	B	8th	34	private

分析单位是教室，但我想把它变成一个数据框，其中分析单位是学校，但要进行一些计算。像这样：

tibble(
  School = c(1, 2, 3, 4),
  City = c("A", "B", "C", "B"),
  N_6th = c(0, 0, 1, 0), # here is the number of grade 6h classrooms in each school 
  N_7th = c(2,1,0,0),
  N_8th = c(0,0,1,1),
  Students_6th = c(0, 0, 25, 0), # here is the number of students in grade 6th from each school (the sum of all 7th grade classrooms from each school)
  Students_7th = c(43, 25, 0, 0), 
  Students_8th = c(0, 0, 28, 34),
  Type_school = c("public", "private", "public", "private")
)

School	City	N_6th	N_7th	N_8th	Students_6th	Students_7th	Students_8th	Type_school
1	A	0	2	0	0	43	0	public
2	B	0	1	0	0	25	0	private
3	C	1	0	1	25	0	28	public
4	B	0	0	1	0	0	34	private

我正在尝试 pivot_wider()，但这不足以满足我的需要。我需要求和每个学校同年级的教室数和每个学校同年级的学生数

Answer 1

按 return 计数和 'Number_Students' 的 sum 进行分组，然后使用 pivot_wider 并将 names_from 指定为 'Grade' 和 values_from 作为列向量

library(dplyr)
library(tidyr)
df1 %>%
  group_by(School, City, Grade, Type_school) %>%
  summarise(N = n(), Students = sum(Number_Students), .groups = 'drop') %>%
  pivot_wider(names_from = Grade, values_from = c(N, Students), values_fill = 0)

-输出

# A tibble: 4 × 9
  School City  Type_school N_7th N_6th N_8th Students_7th Students_6th Students_8th
   <dbl> <chr> <chr>       <int> <int> <int>        <dbl>        <dbl>        <dbl>
1      1 A     public          2     0     0           43            0            0
2      2 B     private         1     0     0           25            0            0
3      3 C     public          0     1     1            0           21           28
4      4 B     private         0     0     1            0            0           34

Answer 2

这是另一种方法：无法与 akrun 的完美方法相提并论，但它包含一些有趣的特征，我们如何获得相同的结果：

library(tidyr)
library(dplyr)

df1 <- df %>% 
  pivot_wider(id_cols = c(School, City, Grade, Type_school),
              names_from = "Grade",
              values_from = "Number_Students",
              values_fn = list(Number_Students = length),
              values_fill = 0,
              names_glue = "N_{Grade}") 

df %>% 
  pivot_wider(id_cols = c(School, City, Grade, Number_Students),
              names_from = Grade,
              values_from = Number_Students,
              values_fn = list(Number_Students = sum),
              names_glue = "Students_{Grade}"
              ) %>% 
  right_join(df1, by=c("School", "City"))

  School City  Students_7th Students_6th Students_8th Type_school N_7th N_6th N_8th
   <dbl> <chr>        <dbl>        <dbl>        <dbl> <chr>       <int> <int> <int>
1      1 A               43           NA           NA public          2     0     0
2      2 B               25           NA           NA private         1     0     0
3      3 C               NA           21           28 public          0     1     1
4      4 B               NA           NA           34 private         0     0     1

我怎样才能更广泛地旋转并转换我的数据框？

How can I pivot wider and transform my data frame?

pivot

r

dataframe

tidyr

wrangle