All the column NA values in a dataframe fill with median values in R-CodePudding

I need to fill the null values of all the numerical columns with each column's median value in a data frame. I did the following code.

median_forNumericalNulls <- function(dataframe){
  
  nums <- unlist(lapply(dataframe, is.numeric))  
  
  df_num <- dataframe[ , nums]
  
  df_num[] <- lapply(df_num, function(x) { 
    x[is.na(x)] <- median(x, na.rm = TRUE)
    x
  })      

  return(dataframe)
  
}

median_forNumericalNulls(A)

A is the parent table, which consists of both numerical as well as categorical variables. How can I replace the columns of 'A' dataframe with the output of the function median_forNumericalNulls?

Is there a better way that we can do the same?

CodePudding user response：

May be we need to change the function to directly subset the columns and updating the columns, instead of creating another object and then updating

median_forNumericalNulls <- function(dataframe){
  
  nums <- unlist(lapply(dataframe, is.numeric))  
  
  df_num <- dataframe[ , nums]
  
  dataframe[nums] <- lapply(dataframe[nums], function(x) { 
    x[is.na(x)] <- median(x, na.rm = TRUE)
    x
  })      
  dataframe
  
}

-testing

A <- median_forNumericalNulls(A)

Also, this can be done in a compact way with na.aggregate though

library(zoo)
A <- na.aggregate(A, FUN = median)

Or using tidyverse

library(dplyr)
A <- A %>%
   mutate(across(where(is.numeric), 
         ~ replace(., is.na(.), median(., na.rm = TRUE))))

CodePudding user response：

Here is another approach how you could do it: Example:

librara(dplyr)
iris1 <- iris %>% 
  select(1, 2, 5)
head(iris1, 10) %>% 
  as_tibble() %>% 
  mutate(across(where(is.numeric), ~ifelse(.<= 3, NA, .))) %>% 
  mutate(across(where(is.numeric), ~ifelse(is.na(.), median(.,na.rm = TRUE), .)))

   Sepal.Length Sepal.Width Species
          <dbl>       <dbl> <fct>  
 1          5.1         3.5 setosa 
 2          4.9         3.4 setosa 
 3          4.7         3.2 setosa 
 4          4.6         3.1 setosa 
 5          5           3.6 setosa 
 6          5.4         3.9 setosa 
 7          4.6         3.4 setosa 
 8          5           3.4 setosa 
 9          4.4         3.4 setosa 
10          4.9         3.1 setosa