How to remove spaces from a fixed-width text file in R

Question

I want to load a text file that looks like the following into R, remove all spaces, then rewrite the file.

The way I'd go about this is with readlines(), removing the spaces in each row, then rewriting.

Is that the best way?

This question is different from just asking how to remove spaces from a string as it involves data framed as rows in a predictable pattern.

Don't read it into R: `system("tr -d \"\" < INPUT > OUTPUT")` — pogibas, Mar 19 '18 at 15:04
Great, idea. Shame this is being downvoted though :( @PoGibas — Union find, Mar 19 '18 at 15:05
Possible duplicate of [How to remove all whitespace from a string?](https://stackoverflow.com/questions/5992082/how-to-remove-all-whitespace-from-a-string) — pogibas, Mar 19 '18 at 15:23

score 1 · Accepted Answer · answered Mar 19 '18 at 15:10

1

I'm not sure exactly what you mean by removing the spaces, but if the data is in the format you describe, then read.table can easily read it.

> tbl <- read.table(text = "
    1 061
    061 1
    0 081
    080 1
    0 061
    060 1
    1 051
    051 1
    0 101
    100 1
")
> tbl
    V1  V2
1    1  61
2   61   1
3    0  81
4   80   1
5    0  61
6   60   1
7    1  51
8   51   1
9    0 101
10 100   1

If you then want to merge the columns so you get the strings without spaces, you simply do this

> no_space <- apply(tbl, 1, function(row) paste0(row, collapse = ""))
> no_space
 [1] "161"  "611"  "081"  "801"  "061"  "601"  "151"  "511" 
 [9] "0101" "1001"

If you want that better formated, cat can do it for you.

> cat(paste0(no_space, collapse = "\n"))
161
611
081
801
061
601
151
511
0101
1001

You can write this to a file using the file argument to cat.

answered Mar 19 '18 at 15:10

Thomas Mailund

1,674
10
16

This looks good but is removing the zeros. Any way to preserve? – Union find Mar 19 '18 at 15:11
1

Absolutely. Add `colClasses = c("character", "character")` as an option to `read.table`. This will prevent R from interpreting the input as numerical, and since we want it as strings anyway, the rest will work. – Thomas Mailund Mar 19 '18 at 15:20
1

Use `do.call` instead of `apply` if speed is important. It's much more efficient. – Roland Mar 19 '18 at 15:50
1

How do you use `do.call` to apply a function over all rows of a table? I don't see an obvious way to replace `apply` by `do.call`... – Thomas Mailund Mar 19 '18 at 16:05
You can use with paste0. `df <- do.call(paste0, df)` @ThomasMailund – Union find Mar 19 '18 at 20:52
Ah! Yes. Clever. – Thomas Mailund Mar 19 '18 at 20:58

How to remove spaces from a fixed-width text file in R

1 Answers1