Creating list of lists with summary statistics for input to summary_table() in R

问题

I am following the instructions laid out here to create a clean table of summary statistics.

In these instructions, the input to the summary_table() function is a list of lists, as shown here:

our_summary1 <-
  list("Miles Per Gallon" =
   list("min" = ~ min(.data$mpg),
        "max" = ~ max(.data$mpg),
        "mean (sd)" = ~ qwraps2::mean_sd(.data$mpg)),
   "Displacement" =
   list("min" = ~ min(.data$disp),
        "median" = ~ median(.data$disp),
        "max" = ~ max(.data$disp),
        "mean (sd)" = ~ qwraps2::mean_sd(.data$disp)),
   "Weight (1000 lbs)" =
   list("min" = ~ min(.data$wt),
        "max" = ~ max(.data$wt),
        "mean (sd)" = ~ qwraps2::mean_sd(.data$wt)),
   "Forward Gears" =
   list("Three" = ~ qwraps2::n_perc0(.data$gear == 3),
        "Four"  = ~ qwraps2::n_perc0(.data$gear == 4),
        "Five"  = ~ qwraps2::n_perc0(.data$gear == 5))
   )

I have 48 variables in my dataset, and each variable has its own column. Is there a cleaner way for me to cycle through all the columns in my dataframe to create an object like the one above without typing it out manually like this? I would ideally prefer a solution using the tidyverse.

One thing I was considering doing was changing my data to long format, then using group_by() to group by each of the columns from the original data, then using summarise(). However, my understanding is that this would yield a single list, not a list of lists like is necessary for summary_table().

If there is a completely different way of creating a summary table than what I am trying to do here, please let me know. This one looked the neatest of the options I was considering. For each variable, I'd like to be able to rename it and include the minimum value, maximum value, mean, and standard deviation.

回答1:

As you noted, you could turn your data to a longer format and use summarize(). The trick is to create a list column within each summarize:

library(dplyr)
library(tidyr)

summarized <- mtcars %>%
  pivot_longer(cols = c(mpg, wt, disp)) %>%
  group_by(name) %>%
  summarize(lst = list(list(mean = mean(value),
                            max = max(value),
                            min = min(value),
                            sd = sd(value))))

summarized
#> # A tibble: 3 x 2
#>   name  lst             
#> * <chr> <list>          
#> 1 disp  <named list [4]>
#> 2 mpg   <named list [4]>
#> 3 wt    <named list [4]>

This can then be turned into a list of lists with deframe() from the tibble package.

library(tibble)
result <- deframe(summarized)

str(result)
#> List of 3
#>  $ disp:List of 4
#>   ..$ mean: num 231
#>   ..$ max : num 472
#>   ..$ min : num 71.1
#>   ..$ sd  : num 124
#>  $ mpg :List of 4
#>   ..$ mean: num 20.1
#>   ..$ max : num 33.9
#>   ..$ min : num 10.4
#>   ..$ sd  : num 6.03
#>  $ wt  :List of 4
#>   ..$ mean: num 3.22
#>   ..$ max : num 5.42
#>   ..$ min : num 1.51
#>   ..$ sd  : num 0.978

回答2:

Nice solution by David (upvoted)! And since I quick typed a SIMPLE solution using lapply, here it is for you to play around with:

library(magrittr)
library(qwraps2)
df = data.frame(matrix(rnorm(40), nrow=10))
df

xlist <- function(x)
  { list(
  "min" = min(x),
  "max" = max(x),
  "mean (sd)" = mean_sd(x))}

attach(df)
Smry <- lapply(1:4,FUN=xlist)
Smry

Just format the list output the way you like it, in the function - the key here is 1) Make a function & 2) Use lapply

来源：https://stackoverflow.com/questions/61580491/creating-list-of-lists-with-summary-statistics-for-input-to-summary-table-in-r

标签

dplyr

apply

tidyverse

summary