Create R objects of all files in a folder via for loop

Question

I have a bunch of repetitive code that I think I can make more efficient by putting into a for loop; been struggling with how to make them objects in R though.

Folder called input has 10 files titled "2010.txt, 2011.txt, ... 2019.txt"

LOOP ONE

files <- list.files("../input")

#Each Year File Path

y2010 <- read_file(glue("../input/", files[1], sep = ""))
y2011 <- read_file(glue("../input/", files[2], sep = ""))
...
y2019 <- read_file(glue("../input/", files[10], sep = ""))

From this I would like to do the following:

##Dataframe of each year's data
all_text <- rbind(y2010,y2011,y2012,y2013,y2014,y2015,y2016,y2017,y2018,y2019)

LOOP TWO Now I would like to take each year and make new "tok201x" objects.

###Each year
tok2010 <- data_frame(text = y2010) %>%
  unnest_tokens(word, text)

tok2011 <- data_frame(text = y2011) %>%
  unnest_tokens(word, text)

...

tok2019 <- data_frame(text = y2019) %>%
  unnest_tokens(word, text)

LOOP THREE Lastly, take the "tok201x" and feed them in to the sentiment code.

#2010
nrc2010 <- tok2010 %>%
  inner_join(get_sentiments("nrc")) %>% # pull out only sentiment words
  count(sentiment) %>% # count each 
  spread(sentiment, n, fill = 0)# made data wide rather than narrow

#2011
nrc2011 <- tok2011 %>%
  inner_join(get_sentiments("nrc")) %>% # pull out only sentiment words
  count(sentiment) %>% # count each 
  spread(sentiment, n, fill = 0)# made data wide rather than narrow

...

#2019
nrc2019 <- tok2019 %>%
  inner_join(get_sentiments("nrc")) %>% # pull out only sentiment words
  count(sentiment) %>% # count each 
  spread(sentiment, n, fill = 0)# made data wide rather than narrow

And have these all stored in a list.

I was playing around with assign() but it was not working out the way I hoped.

EDIT: Using @desval's code with lapply(), I broke the function up. The purpose of this is to combine the lists into one df. How do i accomplish this though?

custom.function1 <- function(x){
  #debug x <- files[1]
  tmp <- read_file(x)
  tmp <- tibble(text = tmp)
return(tmp)
}

custom.function2 <- function(x){
tmp <- tmp %>%
  unnest_tokens(word, text) %>%
  inner_join(get_sentiments("nrc")) %>% # pull out only sentiment words
  count(sentiment) %>% # count each 
  spread(sentiment, n, fill = 0)
return(tmp)
}

out1 <- lapply(files, function1)

##Take all year data and combine into one dataframe, previously...
outYEAR <- matrix(unlist(out1), ncol = 10, byrow = TRUE)
outYEAR <- outYEAR %>% 
    pivot_longer(everything(), names_to = 'year', values_to = 'text') 
##This does not work....

out2 <- lapply(out1, function2)

##Again, combine to one dataframe, previously...
out2YEAR <- matrix(unlist(out2), ncol = 10, byrow = TRUE)
out2YEAR <- out2YEAR %>% 
    pivot_longer(everything(), names_to = 'year', values_to = 'text') 
#THIS DOES NOT WORK.

The collective df's need to be "matrix" not "tbl_df".

Answer 1

I think you might be better off using lapply. I am not sure why it is necessary to read in all the files, rbind them, and the separate them again. If it is not, something along these lines could work:

 library(janeaustenr)
library(tidytext)
library(textdata)
library(tidyverse)
library(data.table)

# some generated data in your directory
d <-  tibble(txt = prideprejudice[1:10])
writeLines(d$txt, "2010.txt")
writeLines(d$txt, "2011.txt")

# list of files
files <- list.files(pattern = "\\d{4}")

custom.function1 <- function(x){
  tmp <- read_file(x)
  tmp <- tibble(text = tmp)
  return(tmp)
}
out1 <- lapply(files, custom.function1)


custom.function2 <- function(x){
  tmp <- x %>% unnest_tokens(word, text) %>%
    inner_join(get_sentiments("nrc")) %>% # pull out only sentiment words
    count(sentiment) %>% # count each 
    spread(sentiment, n, fill = 0)
  tmp <- setDT
  return(tmp)
}
out2 <- lapply(out1, custom.function2)

Now rowbind them (maybe be possible without data.table, but it s very convenient):

out1_all <- out1
out1_all <- lapply(out1_all, setDT) %>% rbindlist(. , id="id_var")

out2_all <- out2
out2_all <- lapply(out2_all, setDT) %>% rbindlist(. , id="id_var")