I have a nested tibble with a column that is basically a bunch of different summary statistics for that group, in the example below it's the same as fun1_result. These were calculated with fun1.
I now want to append a new statistic to the fun1_result column by using fun2. I was guessing I should use map_dbl but I don't actually know what the exact syntax would be to create an additional column within the fun1_result column, basically appending a column to fun1_result.
library(tidyverse)
fun1 = function(df){
result1 = min(df[,"Sepal.Width"])/max(df["Sepal.Length"])
result2 = min(df[,"Sepal.Length"])/max(df["Sepal.Length"])
results = tibble(result1,result2)
return(results)
}
fun2 = function(df){
result3 = min(df[,"Sepal.Width"])/2
return(result3)
}
df1 = iris %>% group_by(Species) %>% nest()
df1 = df1 %>% mutate(fun1_result = map(.x = data,~fun1(.x)))
The result I would be looking for would be the same as the current nested dataframe (after fun1 is applied) with tibbles 1x3 long and an additional statistic as a third column.
I'm sure it's very easy, but I just haven't been able to achieve this as when I used something like this it doesn't work df1 %>% mutate(fun1_result = map_dbl(.x = data, ~fun2(.x)))
Thank you
A way to do it could be to use bind_cols:
library(purrr)
library(dplyr)
df1 %>%
mutate(fun1_result = map(.x = data,~fun1(.x))) |>
mutate(fun1_result = list(bind_cols(fun1_result, tibble(result3 = map(.x = data, ~fun2(.x))))))
But it'll be nicer if fun2 returned a tibble like fun1:
fun2 = function(df) {
result3 = min(df[,"Sepal.Width"])/2
return(tibble(result3))
}
df1 %>%
mutate(fun1_result = map(.x = data,~fun1(.x))) |>
mutate(fun1_result = list(bind_cols(fun1_result, map(.x = data, ~fun2(.x)))))
Output:
# A tibble: 3 × 3
# Groups: Species [3]
Species data fun1_result
<fct> <list> <list>
1 setosa <tibble [50 × 4]> <tibble [1 × 3]>
2 versicolor <tibble [50 × 4]> <tibble [1 × 3]>
3 virginica <tibble [50 × 4]> <tibble [1 × 3]>