How to get the output from the function in a column of a dataframe - R

Question

Hi I have a simple function:

same_picking <- function(cena){
  data_model2$price_model2 <- 0.6 + cena * data_model2$item_SKU + 0.4
}

I would like the output to be rewritten in a column of a data.frame.

currently, because I still did not get the first writing of a function the column is still filled with NAs.. but I would like that after every run of a function the values would be rewriten in theat column.

count_code  sifra   item_SKU    price_model2
281         0421    2           NA
683         0499    5           NA
903         0654    3           NA
7390        0942    3           NA
2778        0796    5           NA
2778        0796    7           NA
7066        0907    83          NA
281         0421    2           NA

I have tried with the comands: data.frame and within... but it got me nowhere.

I would appraciate the help. Andraz

Solution:

same_picking <- function(cena){
      data_model2$price_model2 <<- 0.6 + cena * data_model2$item_SKU + 0.4
}

<<- operator allows you to access the object from the ouside. Very clean :)

The comma will not work data_model1, data_model2$price_model2. Do you have two datasets? Perhaps you need data_model2$price_model2 <- 0.6 + cena * data_model2$item_SKU + 0.4; data_model2 Also it is better to use [ instead of $ — akrun
– akrun, Commented Oct 11, 2016 at 7:36
isn't the return statement already specified? "data_model2$price_model2 <- 0.6 + cena * data_model2$item_SKU + 0.4" — Andraž Poje
– Andraž Poje, Commented Oct 11, 2016 at 7:53
No, you are only creating a column. You have to return the dataset object — akrun
– akrun, Commented Oct 11, 2016 at 8:04

m-dz · Accepted Answer · 2016-10-11 10:52:36Z

1

The simplest way would be to return the df from function:

df <- read.table(
  text = "count_code  sifra   item_SKU    price_model2
          281         0421    2           NA
          683         0499    5           NA
          903         0654    3           NA
          7390        0942    3           NA
          2778        0796    5           NA
          2778        0796    7           NA
          7066        0907    83          NA
          281         0421    2           NA",
  header = TRUE)

head(df, 2)
#   count_code sifra item_SKU price_model2
# 1        281   421        2           NA
# 2        683   499        5           NA

# 1st ---------------------------------------------------------------------

same_picking_1 <- function(df, cena){
  df$price_model2 <- 0.6 + cena * df$item_SKU + 0.4
  return(df)
}

df2 <- same_picking_1(df, 1)

head(df2, 2)
#   count_code sifra item_SKU price_model2
# 1        281   421        2            3
# 2        683   499        5            6

Other options, data.table and dplyr:

same_picking_2 <- function(cena, item_SKU){
  0.6 + cena * df$item_SKU + 0.4
}

# data.table --------------------------------------------------------------

library(data.table)
dt <- data.table(df)
dt[, price_model2 := same_picking_2(1, item_SKU)]

head(dt, 2)
#    count_code sifra item_SKU price_model2
# 1:        281   421        2            3
# 2:        683   499        5            6

# dplyr -------------------------------------------------------------------

library(dplyr)
df3 <- df %>% mutate(price_model2 = same_picking_2(1, item_SKU))

head(df3, 2)
#   count_code sifra item_SKU price_model2
# 1        281   421        2            3
# 2        683   499        5            6

Edit after OP comment:

You can also wrap data.table solution into a function

# data.table --------------------------------------------------------------

library(data.table)

same_picking_2_int <- function(cena, item_SKU){
  0.6 + cena * df$item_SKU + 0.4
}

same_picking_2 <- function(dt, cena){
  dt[, price_model2 := same_picking_2_int(cena, item_SKU)]
}

# Use update by reference
dt <- data.table(df)
head(dt, 2)
same_picking_2(dt, 1)
head(dt, 2)

# Slightly more readable, the same output, also utilizes the update by reference of data.table (see tracemem())
dt <- data.table(df)
tracemem(dt)
head(dt, 2)
dt <- same_picking_2(dt, 1)
head(dt, 2)

edited Oct 11, 2016 at 10:52

answered Oct 11, 2016 at 8:39

m-dz

2,3821 gold badge20 silver badges29 bronze badges

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

3 Comments

Andraž Poje Over a year ago

Thanks for your input. Your solution does return the outcome I was expecting. However I would like to solve the whole process within the function. Because I am expecting to be changing the data a lot and this function should be able to rewrite the data by itself...

Andraž Poje Over a year ago

thanks! but check this out: data_model2$price_model2 <<- 0.6 + cena * data_model2$item_SKU + 0.4 If you use a <<- operator you can access the object inside the dataframe. Very clean solution

m-dz Over a year ago

@AndražPoje, 1. You are on the shortest way to a nice, obfuscated, non-reusable code. 2. Are you sure you will never change the data.frame object (or even it's name)?

Collectives™ on Stack Overflow

How to get the output from the function in a column of a dataframe - R

1 Answer 1

Edit after OP comment:

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Edit after OP comment:

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related