CODEWITHSUNDEEP
r
Ivan Cereghetti
Ivan Cereghetti

Reputation: 300

Apply a conditional function in a nested dataframe

I have a nested dataframe like this:

df<-mpg %>% group_by(manufacturer) %>% nest()

And I have to apply this function to clean some values:

func<-function(drv,cty){
  values<-which(drv=="f"& cty<20)
  cty[values]<-0
  cty}

Only on these manufacturers

manufacturers_vector<-c("audi","chevrolet","jeep")

There is any way I can apply the function only if the manufacturer column in my df matches a value in manufacturers_vector?

Upvotes: 2

Views: 73

Answers (1)

akrun
akrun

Reputation: 887088

We filter the data and then use map to loop over the list 'data'

library(dplyr)
library(purrr)
library(ggplot2)

df2 <- df %>%
   filter(manufacturer %in% manufacturers_vector) %>% 
   mutate(out = map(data,  ~ func(.x$drv, .x$cty)))

-output

df2
# A tibble: 3 x 3
# Groups:   manufacturer [3]
#  manufacturer data               out       
#  <chr>        <list>             <list>    
#1 audi         <tibble [18 × 10]> <dbl [18]>
#2 chevrolet    <tibble [19 × 10]> <dbl [19]>
#3 jeep         <tibble [8 × 10]>  <dbl [8]>

-out column output

df2$out
#[[1]]
# [1]  0 21 20 21  0  0  0 18 16 20 19 15 17 17 15 15 17 16

#[[2]]
# [1] 14 11 14 13 12 16 15 16 15 15 14 11 11 14  0 22  0  0  0

#[[3]]
#[1] 17 15 15 14  9 14 13 11

If we want to keep the original data as such without filter, then use map_if

df %>% 
  mutate(out = map_if(data, .f = ~ func(.x$drv, .x$cty),
     .p = manufacturer %in% manufacturers_vector, .else = ~ NA_real_))

-output

# A tibble: 15 x 3
# Groups:   manufacturer [15]
#   manufacturer data               out       
#   <chr>        <list>             <list>    
# 1 audi         <tibble [18 × 10]> <dbl [18]>
# 2 chevrolet    <tibble [19 × 10]> <dbl [19]>
# 3 dodge        <tibble [37 × 10]> <dbl [1]> 
# 4 ford         <tibble [25 × 10]> <dbl [1]> 
# 5 honda        <tibble [9 × 10]>  <dbl [1]> 
# 6 hyundai      <tibble [14 × 10]> <dbl [1]> 
# 7 jeep         <tibble [8 × 10]>  <dbl [8]> 
# 8 land rover   <tibble [4 × 10]>  <dbl [1]> 
# 9 lincoln      <tibble [3 × 10]>  <dbl [1]> 
#10 mercury      <tibble [4 × 10]>  <dbl [1]> 
#11 nissan       <tibble [13 × 10]> <dbl [1]> 
#12 pontiac      <tibble [5 × 10]>  <dbl [1]> 
#13 subaru       <tibble [14 × 10]> <dbl [1]> 
#14 toyota       <tibble [34 × 10]> <dbl [1]> 
#15 volkswagen   <tibble [27 × 10]> <dbl [1]>

Upvotes: 2

Related Questions