A Package to Make R a Little Nicer

Vignette will usually be updated here first:

When I first started with R, there were a few things that bothered me greatly. While I can’t change dynamic typing, it is possible to do things such as:

  1. String addition, subtraction, multiplication, and division
  2. In-place modifiers (à la +=)
  3. Direct assignments to only NA or regex-matched elements
  4. Comparison operators for between, floating point equality, and more
  5. Extra logical operators to make code more consistent
  6. Make nicer (shorter) conversion functions (int() as opposed to as.integer())
  7. Simple checks for usability (e.g is.bad_for_calcs() or is.os_x64())

The above functionality, I’d found myself manually adding into my R projects to clean up the code. Then me an my colleges thought: ‘that all might actually be useful as a package.’ So now it’s a package on CRAN: roperators (pronounced ’rop-er-ators, not r-operators)

To help introduce you to roperators, I put together some use cases where it’ll make your life easier.

String Arithmetic

One of the most common criticisms lobbed at R by Python people (and their wretched, non-curly-brace-using ilk) is the lack of string arithmetic. In a world without roperators one simply had to deny reality and insist that using a paste function doesn’t look any worse than simply using + to concatenate words.

Happily, using roperators, you can now do this:

require(roperators)
my_string <- 'using infix (%) operators ' %+% 'R can do simple string addition'
print(my_string)
## [1] "using infix (%) operators R can do simple string addition"

You can also use %-% to delete bits of text like so:

my_string %-% 'R can do simple string addition'
## [1] "using infix (%) operators "

If ever need to use string multiplication (like some kind of barbarian), you can use %s*% (%*% was taken already)

my_a <- 'a'
my_a %s*% 3
##     a 
## "aaa"
# If a is an unnamed vector, the original value is saved as the element name(s)
# just to make it easier to undo by my_a <- names(my_a)

And, something you can’t do in Python: string division

# How many times does the letter a appear in the string
'an apple a day keeps the malignant spirit of Steve Jobs at bay' %s/% 'a' 
## a 
## 8

String division also works with regular expressions (it is case sensitive):

# How many times is Steve Jobs or apple mentioned?
'an apple a day keeps the malignant spirit of Steve Jobs at bay' %s/% 'Steve Jobs|apple' 
## Steve Jobs|apple 
##                2

In-Place Modifiers (à la +=)

The lack of operators like += that you’d find in other languages is another common criticism of R. Happily, you have roperators.

Now, at the risk of sounding like one of those infomercials where people struggle with clearly trivial tasks, how many times do you end up doing something like this:

iris_data$Sepal.Length <- iris_data$Sepal.Length + 1

Or worse…

iris_data$Sepal.Length[iris_data$Species == 'setosa'] <- iris_data$Sepal.Length[iris_data$Species == 'setosa'] + 1
# ...which may not even fit on the page. 

Without roperators the trivial code above makes me envy the blind. After all, you’re only adding 1 to some values. So, using the greatest-best package formally called roperators:

iris_data$Sepal.Length %+=% 1

Or

iris_data$Sepal.Length[iris_data$Species == 'setosa'] %+=% 1
# ...which is ike a breath of fresh air

The current in-place modifiers included in roperators are:

The last two are similar depending on whether you want to modify the text or replace it outright. Note that they both take two values c(pattern, replacement):

x <- c("a1b", "b1", "c", "d0")
# Replace digits with the letter x
x %regex=% c("\\d+", "i")
 # x is now c("aib", "bi", "c", "di")
print(x)
## [1] "aib" "bi"  "c"   "di"

Replace Missing Values or Regex matches Directly

The last in-place modifiers are %na<-% which works as you’d expect and %regex<-% which is hopefully intuitive enough. This is useful for all those times you’d otherwise need to do something clunky like df$column[is.na(df$column)] <- 0

x <- c(NA, 1, 2, 3)
x %na<-% 0
print(x)
## [1] 0 1 2 3

And to replace by regex… (as opposed to modifying with %regex=%)

x <- c("aib", "bi", "c", "di")
x %regex<-% c('i', '[redacted]')
print(x)
## [1] "[redacted]" "[redacted]" "c"          "[redacted]"

More Comparisons and Logical Operators

This category of roperators is an answer to all those who cry out for help when what should be simple logical statements are either inconsistent looking or, such as the case with floating point equality, god-awful looking.

When 1 == NA should be FALSE

First up: if(a == b) when a and b are both NA. I get it, an NA doesn’t technically equal another NA, however most of the time they, for all intents and purposes, are the same. The solution is simple:

x <- c(NA, 'foo', 'foo', NA)
y <- c(NA, 'foo', 'bar', 'bar')

x %==% y
## [1]  TRUE  TRUE FALSE FALSE

As opposed to:

x == y
## [1]    NA  TRUE FALSE    NA

Think about how many if statements you’ve had break due to a lack of missing-value equality capability. You can also use %<=% and %>=% to handle missing values instead of <= and >=

When (0.1 + 0.1 + 0.1) == 0.3 should be TRUE (i.e. almost always)

The floating point trap is a particular kind of mongrel. Innocent young statistics students are seldom warned about it, and so they go about, using == thinking that it’ll keep working even when a decimal place is present when in reality, is doesn’t always.

Don’t believe me? Oh, my sweet summer child, try this and despair:

(0.1 * 3) == 0.3  # FALSE
(0.1 * 5) == 0.5  # TRUE
(0.1 * 7) == 0.7  # FALSE
(0.1 * 11) == 1.1 # TRUE

(0.1 * 3) >= 0.3 # TRUE
(0.1 * 3) <= 0.3 # FALSE

If you’re feeling panicked about your old scripts, well, I guess you should be.

Happily, you now have roperators

(0.1 * 3)  %~=% 0.3  # TRUE
(0.1 * 5)  %~=% 0.5  # TRUE
(0.1 * 7)  %~=% 0.7  # TRUE
(0.1 * 11) %~=% 1.1  # TRUE

(0.1 * 3) %>~% 0.3 #TRUE
(0.1 * 3) %<~% 0.3 #TRUE

You could use something like isTRUE(all.equal(0.1 * 3, 0.3)) but that looks disgusting.

isTRUE(all.equal(0.1 * 3, 0.3))  # TRUE
isTRUE(all.equal(0.1 * 5, 0.5))  # TRUE
isTRUE(all.equal(0.1 * 7, 0.7))  # TRUE
isTRUE(all.equal(0.1 * 11, 1.1))  # TRUE


isTRUE(all.equal(0.1 * 3, 0.3)) | ((0.1 * 3) > 0.3)
isTRUE(all.equal(0.1 * 3, 0.3)) | ((0.1 * 3) < 0.3)
# I feel dirty even typing that as an example. 

If you have any sense of style, just use %~=% instead.

When x is between a and b

This is a simple shortcut with two variants for end-exclusive and end-inclusive between. you just need to feed in c(lower_bound, upper_bound)

5 %><% c(1, 10)  # TRUE
## [1] TRUE
1 %><% c(1, 10)  # FALSE
## [1] FALSE
1 %>=<% c(1, 10) # TRUE
## [1] TRUE
# note that due to my simple mindedness, at the time of writing, 5 %><% c(10, 1) is FALSE

Note that %>=<% doesn’t support NA equality testing. If you want a variant that does that in a future version, just let me know.

When you need something else

The last set of logical operators are not in, exclusive or, and all-or-nothing.

Not In %ni% was made because it’s just easier to read than negating an in statement. For example:

!1 %in% c(2,3,4)
## [1] TRUE

Which reads “not 1 in [2, 3, 4]?” which just looks wrong. So, we appropriated from the snake-like language:

1 %ni% c(2,3,4)
## [1] TRUE

Which now reads: “1 not in [2, 3, 4]?” That’s just better looking.

Exclusive Or exists in base R as a function, which makes it look inconsistent, for example:

if((a|b) & xor(y, z))

I know it’s finicky, but the roperators way is a touch more consistent:

if((a|b) & (y %xor% z))

That way both expressions are using an operator rather than one or statement using an operator while the other uses a function.

All or Nothing is for those occasions when you want a and b to either both be TRUE or both be FALSE - for two logical variables it’s probably easier to use a == b, but for expressions it can be cleaner:

if((a*2 == b+2) %aon% (x^2 == y*10))
# Compared to 
if((a*2 == b+2) == (x^2 == y*10)) 
# which takes my brain a little bit more time to read

But, like I said, that’s me personally being finicky.

Shorten type conversions

Fair warning: this part of roperators will, I’m sure, be the source of a lot of hate-mail.

Numeric to factor

One of the ugliest things I see in R code is the infamous x <- as.numeric(as.character(x)) when trying to turn a factor with numeric labels (most of which are the fault of dynamic typing) into a number. I can still recall the rage I felt the first time a factor was converted into its levels rather than its labels when using as.numeric().

The simple solution is just a shorthand: x <- f.as.numeric(x) - just chuck an f in front of it and be done with it.

Shorten as.charater and friends.

I’ll give this one to PyPeople, R’s conversion syntax is cumbersome. That’s why roperators includes:

  • chr() short for as.character()
  • num() short for as.numeric()
  • int() short for as.integer()
  • dbl() short for as.double()
  • chr() short for as.character() (if only str() wasn’t already taken)
  • bool() short for as.logical()

Now things like this:

x <- c('TRUE', 'FALSE', 'TRUE', 'TRUE')

percent_true <- paste0(sum(as.integer(as.logical(x))) / length(x)*100, '%')
print(percent_true)
## [1] "75%"

Can be done like this:

percent_true <- (sum(int(bool(x))) / length(x) * 100) %+% '%'
print(percent_true)
## [1] "75%"

Which is arguably easier on the eyes, especially for people who grew up in other programming languages.

We also added as.class to allow arbitrary conversions in those few moments you find yourself wanting to pipe into a conversion that changes by a variable.

foo <- 204
as.class(foo, "roman")
## [1] CCIV

Add more type checks

Sometimes you just want to know that everything is going to be okay. Rather than running multiple checks. If you wanted to be sure something was going to work in R, you could do something like this:

if(is.atomic(x) & (length(x) >= 1) & !is.na(x) & !is_nan(x) & !is.na(as.numeric(x)) & !is.factor(x) & !is.infinite(x) ){
  ...
}

…Which is fine if you’re happy with people thinking you’re a maniac, Or you could just use roperators like so:

if(!is.bad_for_calcs(x)){
  ...
}

And as a convenience function there’s also any_bad_for_calcs() to save you from any(is.bad_for_calcs((x)) because we’re nice like that.

Beyond that, you’ll also find:

To help with basic checks, and for those times when something should either be a certain class or NULL:

System Checks

Often I want to have my packages know what kind of operating system they’re running on. For example, if I’m writing parallel code, my code needs to know if it’s dealing with a unix-based OS or Windows or which kind of R is running. As such, we added some simplified checks.

Content Checks

For checking if a field has at most 1 or 2 unique values.

Complete Cases Shortcuts

If you’re tired of tryping , na.rm = TRUE we made these functions for you.Basically, just add _cc (complete cases) to a function name and it’ll add na.rm = TRUE for you. They work just like the base functions, only with na.rm = TRUE, similar to paste0() being just paste(..., sep ="")

File Checks

When you need to check that the extension of a file is okay, you can uses these checks. Basically these check the file extensions and for custom cases use check_ext_against().

File Readers

If you work with pipe- or tab-delimited tables, we added read.tsv() and read.psv()

Paste & Cat helpers

Get first, last, nth, most frequent element/word

For basic vectors, it’s pretty intuitive

my_stuff <- c(1:10, 10, 5)

# These are straight forward
get_1st(my_stuff)    # 1
get_nth(my_stuff, 3) # 3
get_last(my_stuff)   # 5

# Returns numeric vector of mode(s) if x is numeric
get_most_frequent(my_stuff) # c(10, 5)

# Else it returns a character vector
my_chars <- c("a", "b", "b", "a", "g", "o", "l", "d")
get_most_frequent(my_chars) # c("a", "b")

# can collapse into a single string (for convienience)
get_most_frequent(my_chars, collapse = " & ") # "a & b"

For pulling apart strings

generic_string <- "Who's A good boy? Winston's a good boy!"

get_1st_word(generic_string)    # Who's
get_nth_word(generic_string, 3) # good
get_last_word(generic_string)   # boy!

# default ignores case and punctuation
get_most_frequent_word(generic_string) # c("a", "boy", "good")

# can change like so:
get_most_frequent_word(generic_string, ignore.case = FALSE, ignore.punct = FALSE) 
# "good"