将 broom::glance 的结果附加在一起

如何解决将 broom::glance 的结果附加在一起

假设我有这个数据：

df <- structure(list(a_bracket = structure(c(9L,8L,9L,9L),.Label = c("0-15","16-20","21-60","61-100","101-500","501-1000","1001-3500","3501-5000","5001+"),class = "factor"),b_bracket = structure(c(3L,2L,3L,4L,1L,4L),.Label = c("18-25","26-35","36-40","41-45","46-48","49-70","71+"),gender = structure(c(2L,2L),.Label = c("Female","Male"),q1 = structure(c(2L,4L
    ),.Label = c("I don't\nlike a thing","I don't\na thing at all","I like a\nthing","Ambivalent about\nthe thing"),q2 = structure(c(3L,1L),.Label = c("Neither like\nnor dislike","Somewhat\ndislike","Somewhat\nlike","Strongly\ndislike","Strongly\nlike"),q3 = structure(c(2L,.Label = c("Moderately","Not at\nall","Quite","Slightly","Very"
    ),class = "factor")),row.names = c(NA,-6L),class = c("tbl_df","tbl","data.frame"))

df

# A tibble: 6 x 6
  a_bracket b_bracket gender q1                            q2                          q3           
  <fct>     <fct>     <fct>  <fct>                         <fct>                       <fct>        
1 5001+     36-40     Male   "I don't\na thing at all"     "Somewhat\nlike"            "Not at\nall"
2 3501-5000 26-35     Male   "I don't\na thing at all"     "Somewhat\ndislike"         "Not at\nall"
3 5001+     36-40     Male   "Ambivalent about\nthe thing" "Neither like\nnor dislike" "Not at\nall"
4 5001+     41-45     Male   "I like a\nthing"             "Neither like\nnor dislike" "Quite"      
5 5001+     18-25     Female "I don't\nlike a thing"       "Strongly\ndislike"         "Not at\nall"
6 5001+     41-45     Male   "Ambivalent about\nthe thing" "Neither like\nnor dislike" "Moderately"

我正在尝试运行一系列模型，提取 r 平方和 AIC，并将它们附加到一个新的 df 中，因变量的名称作为第三行。

这是我的尝试：

model_stats <- function(data){
  
  mod <- glance(
    lm(as.numeric(data) ~ 
         a_bracket + 
         b_bracket + 
         gender,data = df))

  tibble(
    r_squared = mod %>% select(r.squared),AIC = mod %>% select(AIC)
  )
}

map_dfr(
  df %>% 
    select(starts_with("q")),model_stats,.id = "question"
) %>% unnest()

但出于某种原因，我不明白这会根据我正在运行的模型数量将输出重复 N 次。

有谁知道我在这里做错了什么？

解决方法

试试这个 -

library(tidyverse)
library(broom)

model_stats <- function(data){
  
  mod <- glance(
    lm(as.numeric(data) ~ 
         a_bracket + 
         b_bracket + 
         gender,data = df))
  
  tibble(
    r_squared = mod %>% pull(r.squared),AIC = mod %>% pull(AIC)
  )

df %>%
  select(starts_with('q')) %>%
  map_df(model_stats,.id = 'question')

# question r_squared   AIC
#  <chr>        <dbl> <dbl>
#1 q1        6.59e- 1  21.8
#2 q2        7.5 e- 1  20.4
#3 q3        2.22e-31  20.4