如何折叠频率表的行以将其计数添加到新列中？

Question

我有一个带有样本分类的数据框：

 Seq_ID   Family Father   Mother   Sex    Role    Type  
   <chr>     <dbl> <chr>    <chr>    <chr>  <chr>   <chr> 
 1 SSC02219 11000. 0        0        Male   Father  Parent
 2 SSC02217 11000. 0        0        Female Mother  Parent
 3 SSC02254 11000. SSC02219 SSC02217 Male   Proband Child 
 4 SSC02220 11000. SSC02219 SSC02217 Female Sibling Child 
 5 SSC02184 11001. 0        0        Male   Father  Parent
 6 SSC02181 11001. 0        0        Female Mother  Parent
 7 SSC02178 11001. SSC02184 SSC02181 Male   Proband Child 
 8 SSC03092 11002. 0        0        Male   Father  Parent
 9 SSC03078 11002. 0        0        Female Mother  Parent
10 SSC03070 11002. SSC03092 SSC03078 Female Proband Child 

目前，要从 a 转到 b，我必须这样做：

library(tidyverse)
library(janitor)

sample.df %>% tabyl(Role, Sex) %>% 
  adorn_totals(where=c("row", "col") ) %>% 
  as.tibble() %>% select(1,4,3,2) %>%
  # Part 2
  mutate(type=c("parent", "parent", "child", "child", " ")) %>% 
  inner_join(., group_by(., type) %>% 
  summarise(total=sum(Total))) %>% 
  select(5,6,1,2,3,4)

我觉得这是一个非常简单的解决方法。有没有更直接的方法来做 dplyr 的第二部分？

一个

b

Answer 1

这是一个选项。 as.tibble 不是必需的。 mutate 和 case_when 当您有很多类要分配给“父”或“子”时更易于管理。 inner_join 不是必需的，因为我们可以使用 group_by 和 mutate 来计算 total。最后，我喜欢在使用select 函数时记下列名，因为这样我以后会更容易阅读，但是您当然可以使用列索引，只要您确信列索引会是无论您在管道操作中包含哪些新分析，都不会改变。

library(tidyverse)
library(janitor)

sample.df %>% 
  tabyl(Role, Sex) %>% 
  adorn_totals(where=c("row", "col")) %>% 
  select(Role, Total, Male, Female) %>%
  # Part 2
  mutate(type = case_when(
    Role %in% c("Mother", "Father")      ~"parent",
    Role %in% c("Proband", "Sibling")    ~"child",
    TRUE                                 ~" "
  )) %>% 
  group_by(type) %>% 
  mutate(total = sum(Total)) %>%
  ungroup() %>%
  select(type, total, Role, Total, Male, Female)
# # A tibble: 5 x 6
#   type   total Role    Total  Male Female
#   <chr>  <dbl> <chr>   <dbl> <dbl>  <dbl>
# 1 parent    6. Father     3.    3.     0.
# 2 parent    6. Mother     3.    0.     3.
# 3 child     4. Proband    3.    2.     1.
# 4 child     4. Sibling    1.    0.     1.
# 5 " "      10. Total     10.    5.     5.

数据

library(tidyverse)
library(janitor)

sample.df <- read.table(text = "Seq_ID   Family Father   Mother   Sex    Role    Type  
 1 SSC02219 11000  0        0        Male   Father  Parent
 2 SSC02217 11000  0        0        Female Mother  Parent
 3 SSC02254 11000  SSC02219 SSC02217 Male   Proband Child 
 4 SSC02220 11000  SSC02219 SSC02217 Female Sibling Child 
 5 SSC02184 11001  0        0        Male   Father  Parent
 6 SSC02181 11001  0        0        Female Mother  Parent
 7 SSC02178 11001  SSC02184 SSC02181 Male   Proband Child 
 8 SSC03092 11002  0        0        Male   Father  Parent
 9 SSC03078 11002  0        0        Female Mother  Parent
10 SSC03070 11002  SSC03092 SSC03078 Female Proband Child ",
                        header = TRUE, stringsAsFactors = FALSE)

sample.df <- as_tibble(sample.df)

Answer 2

另一种选择是使用knitr

library(janitor)
library(tidyverse)
library(kableExtra)
library(knitr)

sample.df %>% 
  tabyl(Role, Sex) %>%
  adorn_totals(where=c("row", "col")) %>%
  # Part 2
  mutate(type=case_when(
    Role %in% c('Father', 'Mother') ~ 'parent',
    Role %in% c('Proband', 'Sibling') ~ 'child',
    TRUE ~ ''
  )) %>%
  group_by(type) %>%
  mutate(total=sum(Total)) %>%
  ungroup() %>%
  kable("html") %>%
  kable_styling(c("striped", "bordered")) %>%
  collapse_rows(columns = c(5,6))

输出为：

样本数据：

sample.df  <- structure(list(Seq_ID = c("SSC02219", "SSC02217", "SSC02254", 
"SSC02220", "SSC02184", "SSC02181", "SSC02178", "SSC03092", "SSC03078", 
"SSC03070"), Family = c(11000L, 11000L, 11000L, 11000L, 11001L, 
11001L, 11001L, 11002L, 11002L, 11002L), Father = c("0", "0", 
"SSC02219", "SSC02219", "0", "0", "SSC02184", "0", "0", "SSC03092"
), Mother = c("0", "0", "SSC02217", "SSC02217", "0", "0", "SSC02181", 
"0", "0", "SSC03078"), Sex = c("Male", "Female", "Male", "Female", 
"Male", "Female", "Male", "Male", "Female", "Female"), Role = c("Father", 
"Mother", "Proband", "Sibling", "Father", "Mother", "Proband", 
"Father", "Mother", "Proband"), Type = c("Parent", "Parent", 
"Child", "Child", "Parent", "Parent", "Child", "Parent", "Parent", 
"Child")), row.names = c("1", "2", "3", "4", "5", "6", "7", "8", 
"9", "10"), class = c("tbl_df", "tbl", "data.frame"))