r - writing a function that includes for i loop - r

Question

I scanned similar questions previously answered but couldn't find the thread that is specific to my problem.
I have a number of datasets that all have five flagging columns (binary) at the end.
The aim is to produce an output that summarises the specified column in each dataset by each flag.
Hence, each output is a list of five summary tables.
library(tidyverse)
library(janitor)
## mydataset1
mydataset1 <- tibble(id = 1:100,
column_000 = sample(1:16, 100, replace = TRUE),
flag1 = sample(0:1, 100, replace = TRUE),
flag2 = sample(0:1, 100, replace = TRUE),
flag3 = sample(0:1, 100, replace = TRUE),
flag4 = sample(0:1, 100, replace = TRUE),
flag5 = sample(0:1, 100, replace = TRUE))
## summary table function
get_table <- function(data, column) {
data %>%
# select the flag
filter(data[[i]] == 1) %>%
# summary table
tabyl(column) %>%
arrange(desc(n)) %>%
top_n(5, n)
}
## list of tables function
output_list <- function(data, column) {
# empty list
output <- list()
# for loop - go through each flagging column
for (i in (length(data)-4):length(data)) {
output[[i]] <- get_table(data, column)
}
# for some reason, there are NULL list items for all other columns
output <- compact(output)
# rename and print
names(output) <- names(data)[(length(data)-4):length(data)]
print(output)
}
### execute
output_list(mydataset1, "column_000")
# error
### manually executing the function works fine
# empty list
output <- list()
# for loop - go through each flagging column
for (i in (length(mydataset1)-4):length(mydataset1)) {
output[[i]] <- get_table(mydataset1, "column_000")
}
# for some reason, there are NULL list items for all other columns
output <- compact(output)
# rename and print
names(output) <- names(mydataset1)[(length(mydataset1)-4):length(mydataset1)]
print(output)
This is what I have for now.
If I execute the contents of output_list function manually, it works fine.
However, if I execute it as a function, it gives me an error that object i is not found.
Where did I get it wrong? Please help!

Answer 1

Pass i as an input to get_table function.
library(tidyverse)
library(janitor)
get_table <- function(data, column, i) {
data %>%
# select the flag
filter(data[[i]] == 1) %>%
# summary table
tabyl(column) %>%
arrange(desc(n)) %>%
top_n(5, n)
}
Make the corresponding changes in output_list function.
output_list <- function(data, column) {
# empty list
output <- list()
# for loop - go through each flagging column
for (i in (length(data)-4):length(data)) {
output[[i]] <- get_table(data, column, i)
}
# for some reason, there are NULL list items for all other columns
output <- compact(output)
# rename and print
names(output) <- names(data)[(length(data)-4):length(data)]
print(output)
}
Run the function -
output_list(mydataset1, "column_000")

Answer 2

In your get table function, you are using "i" but not declaring "i" in the function argument. Your code works fine when you run code separately because i value gets assigned from the for loop in the global environment. if you intend to use i from for loop in get_table function you can just declare it. See code below.
library(tidyverse)
library(janitor)
## mydataset1
mydataset1 <- tibble(id = 1:100,
column_000 = sample(1:16, 100, replace = TRUE),
flag1 = sample(0:1, 100, replace = TRUE),
flag2 = sample(0:1, 100, replace = TRUE),
flag3 = sample(0:1, 100, replace = TRUE),
flag4 = sample(0:1, 100, replace = TRUE),
flag5 = sample(0:1, 100, replace = TRUE))
## summary table function
get_table <- function(data, column) {
data %>%
# select the flag
filter(data[[i]] == 1) %>%
# summary table
tabyl(column) %>%
arrange(desc(n)) %>%
top_n(5, n)
}
## list of tables function
output_list <- function(data, column) {
# empty list
output <- list()
# for loop - go through each flagging column
for (i in (length(data)-4):length(data)) {
output[[i]] <- get_table(data, column)
}
# for some reason, there are NULL list items for all other columns
output <- compact(output)
# rename and print
names(output) <- names(data)[(length(data)-4):length(data)]
print(output)
}
### execute
output_list(mydataset1, "column_000")
# error
### manually executing the function works fine
# empty list
output <- list()
# for loop - go through each flagging column
for (i in (length(mydataset1)-4):length(mydataset1)) {
output[[i]] <- get_table(mydataset1, "column_000")
}
# for some reason, there are NULL list items for all other columns
output <- compact(output)
# rename and print
names(output) <- names(mydataset1)[(length(mydataset1)-4):length(mydataset1)]
print(output)