问题
Standard R output looks like this
> 3
[1] 3
To remove the prefix 1 you can use
> cat(3)
3
Is there a way to remove this globally? Or do you have to wrap cat() around everything?
Further to that, I'm using this within knitr, so if there isn't an R global setting, there may be a knitr wide setting, I did look, but couldn't see one.
Edit: It was asked why one would want this, something like if you wanted to structure a report like the below. The [1] is just not needed and means nothing to none R users (aka, the audience).
Additional Info: In knitr you can use \Sexpr{} or r ... to evaluate something in line, and in that scenario it doesn't print the [1]. For example if you had:
There are `r sum(mtcars$cyl==6)` cars with 6 cylinders.
You would get:
There are 7 cars with 6 cylinders.
As your output, not:
There are [1] 7 cars with 6 cylinders.
回答1:
Nothing is impossible. Take a look at what can be done with knitr hooks.
Have fun!
Rmarkdown script gist
# A Prefix nulling hook.
# Make sure to keep the default for normal processing.
default_output_hook <- knitr::knit_hooks$get("output")
# Output hooks handle normal R console output.
knitr::knit_hooks$set( output = function(x, options) {
comment <- knitr::opts_current$get("comment")
if( is.na(comment) ) comment <- ""
can_null <- grepl( paste0( comment, "\\s*\\[\\d?\\]" ),
x, perl = TRUE)
do_null <- isTRUE( knitr::opts_current$get("null_prefix") )
if( can_null && do_null ) {
# By default R print output aligns at the right brace.
align_index <- regexpr( "\\]", x )[1] - 1
# Two cases: start or newline
re <- paste0( "^.{", align_index, "}\\]")
rep <- comment
x <- gsub( re, rep, x )
re <- paste0( "\\\n.{", align_index, "}\\]")
rep <- paste0( "\n", comment )
x <- gsub( re, rep, x )
}
default_output_hook( x, options )
})
knitr::opts_template$set("kill_prefix"=list(comment=NA, null_prefix=TRUE))
Normal:
```{r}
print( 1:50 )
```## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
## [24] 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
## [47] 47 48 49 50
Null prefix
```{r, null_prefix=TRUE}
print( 1:50 )
```## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
## 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
## 47 48 49 50
Set as default (and remove comment string):
```{r}
knitr::opts_chunk$set(opts.label="kill_prefix")
```
```{r}
print( 1:50 )
``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
47 48 49 50
Ensure we aren't killing strings with [:digit:] patterns.
```{r}
print( paste0( paste0("[", 1:50), "]" ),quote = FALSE)
```
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
[15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28]
[29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]
[43] [44] [45] [46] [47] [48] [49] [50]
回答2:
Can't be done. But I would hesitate to either do it, or just wrap everything in a cat. Why?
Firstly, R users expect to see the [1] label, and its helpful when the vector is longer than one line, or if its a matrix with rows and column labels. Why do you want to do this?
Secondly, now you know why you want to do this, write a function named after the reason you want to do this, that does this, and wrap your outputs in that function. Because I reckon if you want fine control over the output you are pretty soon going to go "Hmmm cat isn't quite cutting it for me, I want to change this and that and this and that" and that's easier if you have wrapped your output in a function you can change. At first it might just be cleanprint=function(...){cat(...)} but then its easier to move on from there by redefining cleanprint than tweaking cat. You might want to use sprintf, or set the number of decimal digits or something.
来源：https://stackoverflow.com/questions/22524822/how-can-i-remove-the-prefix-index-indicator-1-in-knitr-output