How to apply a self-defined function to a specific column in R and avoid 'longer object length is not a multiple of shorter object length' problem?

Question

Hello and good day to the one reading this!

I am trying to work with gridded data (e.g., lon, lat) and am currently figuring out how to 'round' my values to the closest values in a certain sequence.

For example, I first create a sequence for my gridded longitude values:

lon_seq <- c(seq(120.125, 125.525, 0.05)) # a sequence of even distribution from 120.125 to 125.525

Then, I define this function:

choose_lon <- function(lon_coord){
  lon2 <- lon_seq[which(abs(lon_seq - lon_coord) == min(abs(lon_seq - lon_coord)))][1]
  base::return(lon2)
}

To which if I run with a data I gathered from fieldwork would return a longitude value from the defined grid which is closest to my data (i.e., 123.3729° is closest to 123.375°):

> choose_lon(123.3729)
[1] 123.375

However, if I try to use it in a dataframe, it would return a ! longer object length is not a multiple of shorter object length error. This is a sample data frame and the codes I tried to run:

require(dplyr)

df <- data.frame(
  place = c('A', 'B', 'C', 'D', 'E'),
  code = c('1', '1', '2', '3', '2'),
  lon = c(123.4036, 123.7555, 120.6116, 124.6726, 122.3436)
)

df2 <- df %>%
  dplyr::mutate(lon2 = choose_lon(lon))

The codes above are to hopefully make an output with an additional column of longitude values which have been adjusted accordingly to my "gridded" data.

I have tried to follow the guide to "lengthen" my rows from here but it is to no avail so far.

I hope to kindly ask for your assistance on this matter.

Thank you very much and may you have a nice day!

Edit: I added the require code for the dplyr which is needed.

Answer 1

The same error can be reproduced if you pass two values to your function

choose_lon(c(123.3729, 1233.4036))
#[1] 123.325
Warning messages:

1: In lon_seq - lon_coord :
  longer object length is not a multiple of shorter object length
2: In lon_seq - lon_coord :
  longer object length is not a multiple of shorter object length

So this is what is happening in your code as well. More than one value are passed to the function where your function can only accept one value. To pass only a single value at a time in your function you can add rowwise() to the code.

library(dplyr)

df %>%
  dplyr::rowwise() %>%
  dplyr::mutate(lon2 = choose_lon(lon)) %>%
  data.frame() 

#  place code      lon    lon2
#1     A    1 123.4036 123.425
#2     B    1 123.7555 123.775
#3     C    2 120.6116 120.625
#4     D    3 124.6726 124.675
#5     E    2 122.3436 122.325

Or similarly, with purrr::map_dbl

df %>% dplyr::mutate(lon2 = purrr::map_dbl(lon, choose_lon))