Introducing rlist 0.3

[This article was first published on The blog of Kun Ren, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

rlist 0.3 is released! This package now provides a wide range of functions for dealing with list objects. It can be especially useful when they are used to store non-tabular data.

Two notable features are added in this version. First, and equal() are added in support of fuzzy filtering and searching. Second, List object is added to provide object-based, light-weight chaining operation for list objects.

In the examples, I will use both rlist and pipeR package for easier coding. If you are not yet familiar with either of them, please visit the project pages first.

Fuzzy filtering

Consider the following data in YAML format.

people <- list.parse('
    name: Ken
    age: 24
    interests: [reading, coding]
    friends: [James, Ashley]
    name: James
    age: 23
    interests: [reading, movie, hiking]
    friends: [Ken, David]
    name: Ashley
    age: 25
    interests: [movies, music, reading]
    friends: [Ken, David]
    name: David
    age: 24
    interests: [coding, hiking]
    friends: [Ashley, James]
',type = "yaml")

List of 4
 $ Ken   :List of 4
  ..$ name     : chr "Ken"
  ..$ age      : int 24
  ..$ interests: chr [1:2] "reading" "coding"
  ..$ friends  : chr [1:2] "James" "Ashley"
 $ James :List of 4
  ..$ name     : chr "James"
  ..$ age      : int 23
  ..$ interests: chr [1:3] "reading" "movie" "hiking"
  ..$ friends  : chr [1:2] "Ken" "David"
 $ Ashley:List of 4
  ..$ name     : chr "Ashley"
  ..$ age      : int 25
  ..$ interests: chr [1:3] "movies" "music" "reading"
  ..$ friends  : chr [1:2] "Ken" "David"
 $ David :List of 4
  ..$ name     : chr "David"
  ..$ age      : int 24
  ..$ interests: chr [1:2] "coding" "hiking"
  ..$ friends  : chr [1:2] "Ashley" "James"

In this version, equal() are added to support logical and fuzzy filtering and searching at different levels of exactness. By default, this function tests atomic equality between two atomic vectors unless more parameters are specified. Here are some examples:

Find names of people whose age is 24.

people %>>%
  list.filter(equal(24,age)) %>>%
  list.mapv(name, use.names = FALSE)

[1] "Ken"   "David"

If I set exactly = TRUE then there would be no remaining results since ages are integers in the data but the condition is a numeric value (not exactly an integer parsed in R) unless it is written as 24L.

people %>>%
  list.filter(equal(24,age,exactly = TRUE))

named list()

In fact, equal(exactly = TRUE) calls identical() which tells whether two objects are exactly the same. It not only compares values but their attributes such as names. In this sense, c(1,2) is not identical but atomically equal to c(x=1,y=2). With exactly = TRUE this function becomes the strictest comparer.

In most cases, however, we don't need the target value and original data are exactly the same. Therefore, equal() by default performs atomic equality test.

Sometimes, we need to filter values that includes certain values, here we can specify include = TRUE.

Note that exactly, equal, and include are logical comparers. In addition to them, fuzzy comparers are also supported. They are pattern and dist arguments.

If pattern = TRUE then x serves as a regular expression pattern and equal() tests whether the value matches it. For example, find all people whose interests include something that ends with "ing".

people %>>%
  list.filter(any(equal("ing$",interests,pattern = TRUE))) %>>%
  list.mapv(name, use.names = FALSE)

[1] "Ken"    "James"  "Ashley" "David" 

If dist = is given a number, then it will tolerate all values with a maximum string distance implemented in stringdist package.

Note that in interests, movie and movies both appear but mean the same thing. To tolerate that difference and regard them as equal, specify a string distance in equal(). Now find those who like movies.

people %>>%
  list.filter(any(equal("movies",interests,dist = 1))) %>>%
  list.mapv(name, use.names = FALSE)

[1] "James"  "Ashley"

If there are more records in people and more variants or typos in the term "movies", an appropriate string distance will tolerate them with higher flexibility than regular expressions.

Fuzzy searching

In the new version of rlist, is added to support searching in a list. This function does nothing special but evaluates an expression recursively using rapply.

For example, we search all character vectors which include "James".

people %>>% -> "James" %in% x, classes = "character")

[1] "James"  "Ashley"

[1] "James"

[1] "Ashley" "James" 

It is clearly shown that all character vectors are examined by the condition in contrast with list.filter.

equal() is also designed for facilitating logical and fuzzy searching. For example, search all character vectors that contain values with more than 5 letters and ending with letter "s".

people %>>%"\w{5}s$",pattern = TRUE)),"character")

[1] "movies"  "music"   "reading"

Search all character vectors that contain string like "Kenny" with maximal distance 3.

people %>>%"Kenny", dist = 3)),"character")

[1] "Ken"

[1] "Ken"   "David"

[1] "Ken"   "David"

List environment

Another feature this version introduces is the List object which is designed for light-weight chaining. If you have read my post about pipeR 0.4, you will be familiar with the feature I'm going to talk about.

Here we continue using people data but in List approach. First, let's take a look at the traditional way to use rlist functions to query a list object like people. Suppose we need to extract the names of those who like reading.

people %>>%
  list.filter("reading" %in% interests) %>>%

     Ken    James   Ashley 
   "Ken"  "James" "Ashley" 

Now we can use List object created by List() to make it easier.

  filter("reading" %in% interests)$

$data : character 
     Ken    James   Ashley 
   "Ken"  "James" "Ashley" 

In essence, List is just an environment in which almost all rlist functions are contained but with shorter names. If you need to call external functions, use List()$call(fun,...).

Note that the List environment header in the output indicates that the local functions (or closures) always return the next-level List object to allow chaining. To extract the data in the List object, use [] or $data.


[1] "coding"  "hiking"  "movie"   "movies"  "music"   "reading"

List(people)$cases(friends) []

[1] "Ashley" "David"  "James"  "Ken"   

In both cases the data stored in the object are extracted for further use.

To leave a comment for the author, please follow the link and comment on their blog: The blog of Kun Ren. offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)