In trinker/textmod: Text Cleaning Tools

desc <- suppressWarnings(readLines("DESCRIPTION"))
regex <- "(^Version:\\s+)(\\d+\\.\\d+\\.\\d+)"
loc <- grep(regex, desc)
ver <- gsub(regex, "\\2", desc[loc])
pacman::p_load_current_gh('trinker/numform')
verbadge <- sprintf('<a href="https://img.shields.io/badge/Version-%s-orange.svg"><img src="https://img.shields.io/badge/Version-%s-orange.svg" alt="Version"/></a></p>', ver, ver)
verbadge <- ""

r verbadge

textclean is a collection of tools to clean and normalize text. Many of these tools have been taken from the qdap package and revamped to be more intuitive, better named, and faster. Tools are geared at checking for substrings that are not optimal for analysis and replacing or removing them (normalizing) with more analysis friendly substrings (see Sproat, Black, Chen, Kumar, Ostendorf, & Richards, 2001, doi:10.1006/csla.2001.0169) or extracting them into new variables. For example, emoticons are often used in text but not always easily handled by analysis algorithms. The replace_emoticon() function replaces emoticons with word equivalents.

Other R packages provide some of the same functionality (e.g., english, gsubfn, mgsub, stringi, stringr, qdapRegex). textclean differs from these packages in that it is designed to handle all of the common cleaning and normalization tasks with a single, consistent, pre-configured toolset (note that textclean uses many of these terrific packages as a backend). This means that the researcher spends less time on munging, leading to quicker analysis. This package is meant to be used jointly with the textshape package, which provides text extraction and reshaping functionality. textclean works well with the qdapRegex package which provides tooling for substring substitution and extraction of pre-canned regular expressions. In addition, the functions of textclean are designed to work within the piping of the tidyverse framework by consistently using the first argument of functions as the data source. The textclean subbing and replacement tools are particularly effective within a dplyr::mutate statement.

Functions

The main functions, task category, & descriptions are summarized in the table below:

| Function | Task | Description | |---------------------------|-------------|---------------------------------------| | mgsub | subbing | Multiple gsub | | fgsub | subbing | Functional matching replacement gsub | | sub_holder | subbing | Hold a value prior to a strip | | swap | subbing | Simultaneously swap patterns 1 & 2 | | strip | deletion | Remove all non word characters | | drop_empty_row | filter rows | Remove empty rows | | drop_row/keep_row | filter rows | Filter rows matching a regex | | drop_NA | filter rows | Remove NA text rows | | drop_element/keep_element | filter elements | Filter matching elements from a vector | | match_tokens | filter elements | Filter out tokens from strings that match a regex criteria | | replace_contractions | replacement | Replace contractions with both words | | replace_date | replacement | Replace dates | | replace_email | replacement | Replace emails | | replace_emoji | replacement | Replace emojis with word equivalent or unique identifier | | replace_emoticon | replacement | Replace emoticons with word equivalent | | replace_grade | replacement | Replace grades (e.g., "A+") with word equivalent | | replace_hash | replacement | Replace Twitter style hash tags (e.g., #rstats) | | replace_html | replacement | Replace HTML tags and symbols | | replace_incomplete | replacement | Replace incomplete sentence end-marks | | replace_internet_slang | replacement | Replace Internet slang with word equivalents | | replace_kern | replacement | Replace spaces for >2 letter, all cap, words containing spaces in between letters | | replace_misspelling | replacement | Replace misspelled words with their most likely replacement | | replace_money | replacement | Replace money in the form of $\d+.?\d{0,2} | | replace_names | replacement | Replace common first/last names | | replace_non_ascii | replacement | Replace non-ASCII with equivalent or remove | | replace_number | replacement | Replace common numbers | | replace_ordinal | replacement | Replace common ordinal number form | | replace_rating | replacement | Replace ratings (e.g., "10 out of 10", "3 stars") with word equivalent | | replace_symbol | replacement | Replace common symbols | | replace_tag | replacement | Replace Twitter style handle tag (e.g., @trinker) | | replace_time | replacement | Replace time stamps | | replace_to/replace_from | replacement | Remove from/to begin/end of string to/from a character(s) | | replace_tokens | replacement | Remove or replace a vector of tokens with a single value | | replace_url | replacement | Replace URLs | | replace_white | replacement | Replace regex white space characters | | replace_word_elongation | replacement | Replace word elongations with shortened form | | add_comma_space | replacement | Replace non-space after comma | | add_missing_endmark | replacement | Replace missing endmarks with desired symbol | | make_plural | replacement | Add plural endings to singular noun forms | | check_text | check | Text report of potential issues | | has_endmark | check | Check if an element has an end-mark |

Installation

To download the development version of textclean:

Download the zip ball or tar ball, decompress and run R CMD INSTALL on it, or use the pacman package to install the development version:

if (!require("pacman")) install.packages("pacman")
pacman::p_load_gh(
    "trinker/lexicon",    
    "trinker/textclean"
)

Contact

You are welcome to:
- submit suggestions and bug-reports at: https://github.com/trinker/textclean/issues

Contributing

Contributions are welcome from anyone subject to the following rules:

• Abide by the code of conduct.
• Follow the style conventions of the package (indentation, function & argument naming, commenting, etc.)
• All contributions must be consistent with the package license (GPL-2)
• Submit contributions as a pull request. Clearly state what the changes are and try to keep the number of changes per pull request as low as possible.
• If you make big changes, add your name to the 'Author' field.

Demonstration

Load the Packages/Data

library(dplyr)
library(textshape)
library(lexicon)
library(textclean)

Check Text

One of the most useful tools in textclean is check_text which scans text variables and reports potential problems. Not all potential problems are definite problems for analysis but the report provides an overview of what may need further preparation. The report also provides suggested functions for the reported problems. The report provides information on the following:

cat(sapply(textclean:::.checks, function(x){

    frame <- ifelse(x$is_meta, "1. **%s** - Text variable that %s", "1. **%s** - Text elements that %s")
    sprintf(frame, x$fun, x$problem)

}), sep = '\n')

Note that check_text is running multiple checks and may be slower on larger texts. The user may provide a sample of text for review or use the checks argument to specify the exact checks to conduct and thus limit the compute time.

Here is an example:

x <- c("i like", "<p>i want. </p>. thet them ther .", "I am ! that|", "", NA, 
    "&quot;they&quot; they,were there", ".", "   ", "?", "3;", "I like goud eggs!", 
    "bi\xdfchen Z\xfcrcher", "i 4like...", "\\tgreat",  "She said \"yes\"")
Encoding(x) <- "latin1"
x <- as.factor(x)
check_text(x)

And if all is well the user should be greeted by a cow:

y <- c("A valid sentence.", "yet another!")
check_text(y)

Row Filtering

It is useful to drop/remove empty rows or unwanted rows (for example the researcher dialogue from a transcript). The drop_empty_row & drop_row do empty row do just this. First I'll demo the removal of empty rows.

## create a data set wit empty rows
(dat <- rbind.data.frame(DATA[, c(1, 4)], matrix(rep(" ", 4), 
    ncol =2, dimnames=list(12:13, colnames(DATA)[c(1, 4)]))))

drop_empty_row(dat)

Next we drop out rows. The drop_row function takes a data set, a column (named or numeric position) and regex terms to search for. The terms argument takes regex(es) allowing for partial matching. terms is case sensitive but can be changed via the ignore.case argument.

drop_row(dataframe = DATA, column = "person", terms = c("sam", "greg"))
drop_row(DATA, 1, c("sam", "greg"))
keep_row(DATA, 1, c("sam", "greg"))
drop_row(DATA, "state", c("Comp"))
drop_row(DATA, "state", c("I "))
drop_row(DATA, "state", c("you"), ignore.case = TRUE)

Stripping

Often it is useful to remove all non relevant symbols and case from a text (letters, spaces, and apostrophes are retained). The strip function accomplishes this. The char.keep argument allows the user to retain characters.

strip(DATA$state)
strip(DATA$state, apostrophe.remove = TRUE)
strip(DATA$state, char.keep = c("?", "."))

Subbing

Multiple Subs

gsub is a great tool but often the user wants to replace a vector of elements with another vector. mgsub allows for a vector of patterns and replacements. Note that the first argument of mgsub is the data, not the pattern as is standard with base R's gsub. This allows mgsub to be used in a magrittr pipeline more easily. Also note that by default fixed = TRUE. This means the search pattern is not a regex per-se. This makes the replacement much faster when a regex search is not needed. mgsub also reorders the patterns to ensure patterns contained within patterns don't over write the longer pattern. For example if the pattern c('i', 'it') is given the longer 'it' is replaced first (though order.pattern = FALSE can be used to negate this feature).

mgsub(DATA$state, c("it's", "I'm"), c("<<it is>>", "<<I am>>"))
mgsub(DATA$state, "[[:punct:]]", "<<PUNCT>>", fixed = FALSE)
mgsub(DATA$state, c("i", "it"), c("<<I>>", "[[IT]]"))
mgsub(DATA$state, c("i", "it"), c("<<I>>", "[[IT]]"), order.pattern = FALSE)

Safe Substitutions

The default behavior of mgsub is optimized for speed. This means that it is very fast at multiple substitutions and in most cases works efficiently. However, it is not what Mark Ewing describes as "safe" substitution. In his vignette for the mgsub package, Mark defines "safe" as:

1. Longer matches are preferred over shorter matches for substitution first
2. No placeholders are used so accidental string collisions don't occur

Because safety is sometimes required, textclean::mgsub provides a safe argument using the mgsub package as the backend. In addition to the safe argument the mgsub_regex_safe function is available to make the usage more explicit. The safe mode comes at the cost of speed.

x <- "Dopazamine is a fake chemical"
pattern <- c("dopazamin", "do.*ne")
replacement <- c("freakout", "metazamine")

## Unsafe
mgsub(x, pattern, replacement, ignore.case=TRUE, fixed = FALSE)
## Safe
mgsub(x, pattern, replacement, ignore.case=TRUE, fixed = FALSE, safe = TRUE)
## Or alternatively
mgsub_regex_safe(x, pattern, replacement, ignore.case=TRUE)

x <- "hey, how are you?"
pattern <- c("hey", "how", "are", "you")
replacement <- c("how", "are", "you", "hey")

## Unsafe
mgsub(x, pattern,replacement)
## Safe
mgsub_regex_safe(x, pattern,replacement)

Match, Extract, Operate, Replacement Subs

Again, gsub is a great tool but sometimes the user wants to match a pattern, extract that pattern, operate a function over that pattern, and then replace the original match. The fgsub function allows the user to perform this operation. It is a stripped down version of gsubfn from the gsubfn package. For more versatile needs please see the gsubfn package.

In this example the regex looks for words that contain a lower case letter followed by the same letter at least 2 more times. It then extracts these words, splits them appart into letters, reverses the string, pastes them back together, wraps them with double angle braces, and then puts them back at the original locations.

fgsub(
    x = c(NA, 'df dft sdf', 'sd fdggg sd dfhhh d', 'ddd'),
    pattern = "\\b\\w*([a-z])(\\1{2,})\\w*\\b",
    fun = function(x) {paste0('<<', paste(rev(strsplit(x, '')[[1]]), collapse =''), '>>')}
)

In this example we extract numbers, strip out non-digits, coerce them to numeric, cut them in half, round up to the closest integer, add the commas back, and replace back into the original locations.

fgsub(
    x = c(NA, 'I want 32 grapes', 'he wants 4 ice creams', 'they want 1,234,567 dollars'),
    pattern = "[\\d,]+",
    fun = function(x) {prettyNum(ceiling(as.numeric(gsub('[^0-9]', '', x))/2), big.mark = ',')}
)

Stashing Character Pre-Sub

There are times the user may want to stash a set of characters before subbing out and then return the stashed characters. An example of this is when a researcher wants to remove punctuation but not emoticons. The subholder function provides tooling to stash the emoticons, allow a punctuation stripping, and then return the emoticons. First I'll create some fake text data with emoticons, then stash the emoticons (using a unique text key to hold their place), then I'll strip out the punctuation, and last put the stashed emoticons back.

(fake_dat <- paste(hash_emoticons[1:11, 1, with=FALSE][[1]], DATA$state))
(m <- sub_holder(fake_dat, hash_emoticons[[1]]))
(m_stripped <-strip(m$output))
m$unhold(m_stripped)

Of course with clever regexes you can achieve the same thing:

ord_emos <- hash_emoticons[[1]][order(nchar(hash_emoticons[[1]]))]

## This step ensures that longer strings are matched first but can 
## fail in cases that use quantifiers.  These can appear short but in
## reality can match long strings and would be ordered last in the 
## replacement, meaning that the shorter regex took precedent.
emos <- paste(
    gsub('([().\\|[{}^$*+?])', '\\\\\\1', ord_emos),
    collapse = '|'
)

gsub(
    sprintf('(%s)(*SKIP)(*FAIL)|[^\'[:^punct:]]', emos), 
    '', 
    fake_dat, 
    perl = TRUE
)

The pure regex approach can be a bit trickier (less safe) and more difficult to reason about. It also relies on the less general (*SKIP)(*FAIL) backtracking control verbs that are only implemented in a few applications like Perl & PCRE. Still, it's nice to see an alternative regex approach for comparison.

Replacement

textclean contains tools to replace substrings within text with other substrings that may be easier to analyze. This section outlines the uses of these tools.

Contractions

Some analysis techniques require contractions to be replaced with their multi-word forms (e.g., "I'll" -> "I will"). replace_contrction provides this functionality.

x <- c("Mr. Jones isn't going.",  
    "Check it out what's going on.",
    "He's here but didn't go.",
    "the robot at t.s. wasn't nice", 
    "he'd like it if i'd go away")

replace_contraction(x)

Dates

x <- c(NA, '11-16-1980 and 11/16/1980', "and 2017-02-08 but then there's 2/8/2017 too")

replace_date(x)
replace_date(x, replacement = '<<DATE>>')

Emojis

Similar to emoticons, emoji tokens may be ignored if they are not in a computer readable form. replace_emoji replaces emojis with their word forms equivalents.

x <- read.delim(system.file("docs/r_tweets.txt", package = "textclean"), 
    stringsAsFactors = FALSE)[[3]][1:3]

x

replace_emoji(x)

Emoticons

Some analysis techniques examine words, meaning emoticons may be ignored. replace_emoticon replaces emoticons with their word forms equivalents.

x <- c(
    "text from: http://www.webopedia.com/quick_ref/textmessageabbreviations_02.asp",
    "... understanding what different characters used in smiley faces mean:",
    "The close bracket represents a sideways smile  )",
    "Add in the colon and you have sideways eyes   :",
    "Put them together to make a smiley face  :)",
    "Use the dash -  to add a nose   :-)",
    "Change the colon to a semi-colon ; and you have a winking face ;)  with a nose  ;-)",
    "Put a zero 0 (halo) on top and now you have a winking, smiling angel 0;) with a nose 0;-)",
    "Use the letter 8 in place of the colon for sunglasses 8-)",
    "Use the open bracket ( to turn the smile into a frown  :-("
)

replace_emoticon(x)

Grades

In analysis where grades may be discussed it may be useful to convert the letter forms into word meanings. The replace_grade can be used for this task.

text <- c(
    "I give an A+",
    "He deserves an F",
    "It's C+ work",
    "A poor example deserves a C!"
)
replace_grade(text)

HTML

Sometimes HTML tags and symbols stick around like pesky gnats. The replace_html function makes light work of them.

x <- c(
    "<bold>Random</bold> text with symbols: &nbsp; &lt; &gt; &amp; &quot; &apos;",
    "<p>More text</p> &cent; &pound; &yen; &euro; &copy; &reg;"
)

replace_html(x)

Incomplete Sentences

Sometimes an incomplete sentence is denoted with multiple end marks or no punctuation at all. replace_incomplete standardizes these sentences with a pipe (|) endmark (or one of the user's choice).

x <- c("the...",  "I.?", "you.", "threw..", "we?")
replace_incomplete(x)
replace_incomplete(x, '...')

Internet Slang

Often in informal written and spoken communication (e.g., Twitter, texting, Facebook, etc.) people use Internet slang, shorter abbreviations and acronyms, to replace longer word sequences. These replacements may obfuscate the meaning when the machine attempts to analyze the text. The replace_internet_slang function replaces the slang with longer word equivalents that are more easily analyzed by machines.

x <- c(
    "TGIF and a big w00t!  The weekend is GR8!",
    "NP it was my pleasure: EOM",
    'w8...this n00b needs me to say LMGTFY...lol.',
    NA
)

replace_internet_slang(x)

Kerning

In typography kerning is the adjustment of spacing. Often, in informal writing, adding manual spaces (a form of kerning) coupled with all capital letters is used for emphasis (e.g., "She's the B O M B!"). These word forms would look like noise in most analysis and would likely be removed as a stopword when in fact they likely carry a great deal of meaning. The replace_kern function looks for 3 or more consecutive capital letters with spaces in between and removes the spaces.

x <- c(
    "Welcome to A I: the best W O R L D!",
    "Hi I R is the B O M B for sure: we A G R E E indeed.",
    "A sort C A T indeed!",
    NA
)

replace_kern(x)

Money

There are times one may want to replace money mentions with text or normalized versions. The replace_money function is designed to complete this task.

x <- c(NA, '$3.16 into "three dollars, sixteen cents"', "-$20,333.18 too", 'fff')

replace_money(x)

replace_money(x, replacement = '<<MONEY>>')

Names

Often one will want to standardize text by removing first and last names. The replace_names function quickly removes/replaces common first and last names. This can be made more targeted by feeding a vector of names extracted via a named entity extractor.

x <- c(
    "Mary Smith is not here",
     "Karen is not a nice person",
     "Will will do it",
    NA
)

replace_names(x)
replace_names(x, replacement = '<<NAME>>')

Non-ASCII Characters

R can choke on non-ASCII characters. They can be re-encoded but the new encoding may lack interpretability (e.g., ¢ may be converted to \xA2 which is not easily understood or likely to be matched in a hash look up). replace_non_ascii attempts to replace common non-ASCII characters with a text representation (e.g., ¢ becomes "cent") Non recognized non-ASCII characters are simply removed (unless remove.nonconverted = FALSE).

x <- c(
    "Hello World", "6 Ekstr\xf8m", "J\xf6reskog", "bi\xdfchen Z\xfcrcher",
    'This is a \xA9 but not a \xAE', '6 \xF7 2 = 3', 'fractions \xBC, \xBD, \xBE',
    'cows go \xB5', '30\xA2'
)
Encoding(x) <- "latin1"
x

replace_non_ascii(x)
replace_non_ascii(x, remove.nonconverted = FALSE)

Numbers

Some analysis requires numbers to be converted to text form. replace_number attempts to perform this task. replace_number handles comma separated numbers as well.

x <- c("I like 346,457 ice cream cones.", "They are 99 percent good")
y <- c("I like 346457 ice cream cones.", "They are 99 percent good")
replace_number(x)
replace_number(y)
replace_number(x, num.paste = TRUE)
replace_number(x, remove=TRUE)

Ratings

Some texts use ratings to convey satisfaction with a particular object. The replace_rating function replaces the more abstract rating with word equivalents.

x <- c("This place receives 5 stars for their APPETIZERS!!!",
     "Four stars for the food & the guy in the blue shirt for his great vibe!",
     "10 out of 10 for both the movie and trilogy.",
     "* Both the Hot & Sour & the Egg Flower Soups were absolutely 5 Stars!",
     "For service, I give them no stars.", "This place deserves no stars.",
     "10 out of 10 stars.",
     "My rating: just 3 out of 10.",
     "If there were zero stars I would give it zero stars.",
     "Rating: 1 out of 10.",
     "I gave it 5 stars because of the sound quality.",
     "If it were possible to give them 0/10, they'd have it."
)

replace_rating(x)

Ordinal Numbers

Again, some analysis requires numbers, including ordinal numbers, to be converted to text form. replace_ordinal attempts to perform this task for ordinal number 1-100 (i.e., 1st - 100th).

x <- c(
    "I like the 1st one not the 22nd one.", 
    "For the 100th time stop those 3 things!",
    "I like the 3rd 1 not the 12th 1."
)
replace_ordinal(x)
replace_ordinal(x, TRUE)
replace_ordinal(x, remove = TRUE)
replace_number(replace_ordinal(x))

Symbols

Text often contains short-hand representations of words/phrases. These symbols may contain analyzable information but in the symbolic form they cannot be parsed. The replace_symbol function attempts to replace the symbols c("$", "%", "#", "@", "& "w/") with their word equivalents.

x <- c("I am @ Jon's & Jim's w/ Marry", 
    "I owe $41 for food", 
    "two is 10% of a #"
)
replace_symbol(x)

Time Stamps

Often times the researcher will want to replace times with a text or normalized version. The replace_time function works well for this task. Notice that replacement takes a function that can operate on the extracted pattern.

x <- c(
    NA, '12:47 to "twelve forty-seven" and also 8:35:02',
    'what about 14:24.5', 'And then 99:99:99?'
)

## Textual: Word version
replace_time(x)

## Normalization: <<TIME>>
replace_time(x, replacement = '<<TIME>>')

## Normalization: hh:mm:ss or hh:mm
replace_time(x, replacement = function(y){
        z <- unlist(strsplit(y, '[:.]'))
        z[1] <- 'hh'
        z[2] <- 'mm'
        if(!is.na(z[3])) z[3] <- 'ss'
        textclean::glue_collapse(z, ':')
    }
)

## Textual: Word version (forced seconds)
replace_time(x, replacement = function(y){
        z <- replace_number(unlist(strsplit(y, '[:.]')))
        z[3] <- paste0('and ', ifelse(is.na(z[3]), '0', z[3]), ' seconds')
        paste(z, collapse = ' ')
    }
)

Tokens

Often an analysis requires converting tokens of a certain type into a common form or removing them entirely. The mgsub function can do this task, however it is regex based and time consuming when the number of tokens to replace is large. For example, one may want to replace all proper nouns that are first names with the word name. The replace_token provides a fast way to replace a group of tokens with a single replacement.

This example shows a use case for replace_token:

## Set Up the Tokens to Replace
nms <- gsub("(^.)(.*)", "\\U\\1\\L\\2", lexicon::common_names, perl = TRUE)
head(nms)

## Set Up the Data
x <- textshape::split_portion(sample(c(sample(lexicon::grady_augmented, 20000), 
    sample(nms, 10000, TRUE))), n.words = 12)
x$text.var <- paste0(x$text.var, sample(c('.', '!', '?'), length(x$text.var), TRUE))
head(x$text.var)

head(replace_tokens(x$text.var, nms, 'NAME'))

This demonstration shows how fast token replacement can be with replace_token:

## mgsub
tic <- Sys.time()
head(mgsub(x$text.var, nms, "NAME"))
(toc <- Sys.time() - tic)

## replace_tokens
tic <- Sys.time()
head(replace_tokens(x$text.var, nms, "NAME"))
(toc <- Sys.time() - tic)

tic <- Sys.time()
out <- replace_tokens(rep(x$text.var, 20), nms, "NAME")
toc <- Sys.time() - tic

Now let's amp it up with 20x more text data. That's r f_comma(length(x$text.var) * 20) rows of text (r f_comma(sum(stringi::stri_count_words(x$text.var))*20) words) and r f_comma(length(nms)) replacement tokens in r round(toc, 1) seconds.

tic <- Sys.time()
out <- replace_tokens(rep(x$text.var, 20), nms, "NAME")
(toc <- Sys.time() - tic)

toc

White Space

Regex white space characters (e.g., \n, \t, \r) matched by \s may impede analysis. These can be replaced with a single space " " via the replace_white function.

x <- "I go \r
    to   the \tnext line"
x
cat(x)
replace_white(x)

Word Elongation

In informal writing people may use a form of text embellishment to emphasize or alter word meanings called elongation (a.k.a. "word lengthening"). For example, the use of "Whyyyyy" conveys frustration. Other times the usage may be to be more sexy (e.g., "Heyyyy there"). Other times it may be used for emphasis (e.g., "This is so gooood").

The replace_word_elongation function replaces these un-normalized forms with the most likely normalized form. The impart.meaning argument can replace a short list of known elongations with semantic replacements.

x <- c('look', 'noooooo!', 'real coooool!', "it's sooo goooood", 'fsdfds',
    'fdddf', 'as', "aaaahahahahaha", "aabbccxccbbaa", 'I said heyyy!',
    "I'm liiiike whyyyyy me?", "Wwwhhatttt!", NA)

replace_word_elongation(x)
replace_word_elongation(x, impart.meaning = TRUE)

trinker/textmod documentation built on Nov. 3, 2021, 7:20 p.m.