Function works on dataframe, but gives error when using lapply

Question

I am trying to apply a function across a number of dataframes using lapply. The function works when I invoke it on each of the dataframes individually, but lapply throws an error. The error doesn't seem relevant. I can't work out what the issue is. Here is an example:

a <- data.frame('country' = factor(c(rep(1, 5), rep(2, 5))), 
           'variable' = factor(c(rep('A', 5), rep('B', 5))), 
           'value' = runif(10, 0, 1), 
           'year' = runif(10, 0, 1))

slope <- function(dat) {
  dat %>%
  filter(!value %in% c(-66, -77, -88) & !is.na(value)) %>%
  group_by(country, variable) %>%
  do(data.frame(slope = coef(lm(value ~ year, .))[2])) %>%
  ungroup()
}

This function works:

> slope(a)
    # A tibble: 2 x 3
      country variable  slope
      <fct>   <fct>     <dbl>
    1 1       A         0.140
    2 2       B        -0.150

But lapply doesn't:

   > lapply(a, slope)
     Error in UseMethod("filter_") : 
      no applicable method for 'filter_' applied to an object of class "factor"

I don't understand the error because value, which is filtered, is numeric (not a factor).

> str(a)
'data.frame':   10 obs. of  4 variables:
 $ country : Factor w/ 2 levels "1","2": 1 1 1 1 1 2 2 2 2 2
 $ variable: Factor w/ 2 levels "A","B": 1 1 1 1 1 2 2 2 2 2
 $ value   : num  0.884 0.513 0.835 0.83 0.694 ...
 $ year    : num  0.4288 0.2874 0.0531 0.7793 0.0496 ...

Obviously when using lapply in practice, I would be using it on a number of dataframes. I don't think it makes a difference in the example - i get the same error when trying to do this on a number of dataframes. I assume I am missing something obvious.

when you loop over a, it is looping through the columns, i.e. a vector and is not a data.frame — akrun
– akrun, Commented Dec 15, 2019 at 22:34
@akrun, thanks for this. Seeing your first comment, made me realize that I had used lapply on c('a', 'b', 'c'), rather than on list('a', 'b', 'c') - where 'a', 'b' and 'c' are all dataframes. When I do that it all works. Thanks for this. — MorrisseyJ
– MorrisseyJ, Commented Dec 15, 2019 at 22:43

akrun · Accepted Answer · 2019-12-15 22:41:02Z

1

The issue is that applying lapply on the data.frame, loops through the columns as column is a unit in a data.frame i.e. the output is a list of vectors and it is not a data.frame while the slope function expects a data.frame with columns to act upon.

Also, the OP mentioned about applying the function on a number of data.frames. In that case, place the datasets in a list and apply with lapply i.e.

list(a, a) %>%
   lapply(slope)

Or with a single dataset, wrap with list

list(a) %>%
   lapply(slope)

Or in tidyverse

library(purrr)
list(a) %>%
    map(slope)

answered Dec 15, 2019 at 22:41

akrun

891k38 gold badges590 silver badges700 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

user5249203 Over a year ago

can we pass name columns to the slope function ? If so how ?

Collectives™ on Stack Overflow

Function works on dataframe, but gives error when using lapply

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related