quanteda v0.99 Important Changes
Many important changes to the package have been underway, including API changes, as we approach a stable API, "1.0" release in October 2017. Version 0.99 represents the last version that will contain many of the deprecated object types and methods that date several releases.
v0.99 also implements many enhancements and performance improvements over previous releases. See NEWS.md for details, and Quanteda Structure and Design for a description of the package's underlying logic and design philosophy.
About the package
An R package for managing and analyzing text, created by Kenneth Benoit in collaboration with a team of core contributors: Kohei Watanabe, Paul Nulty, Adam Obeng, Haiyan Wang, Ben Lauderdale, and Will Lowe. Supported by the European Research Council grant ERC-2011-StG 283794-QUANTESS.
For more details, see the package website.
How to cite the package:
To cite package 'quanteda' in publications please use the
following:
Benoit, Kenneth et. al. (). "quanteda: Quantitative Analysis of
Textual Data". R package version: 0.99.12. http://quanteda.io.
A BibTeX entry for LaTeX users is
@Manual{,
title = {quanteda: Quantitative Analysis of Textual Data},
author = {Kenneth Benoit and Kohei Watanabe and Paul Nulty and Adam Obeng and Haiyan Wang and Benjamin Lauderdale and Will Lowe},
note = {R package version 0.99.12},
url = {http://quanteda.io},
}Leave feedback
If you like quanteda, please consider leaving feedback or a testimonial here.
Features
Powerful text analytics
Generalized, flexible corpus management. quanteda provides a comprehensive workflow and ecosystem for the management, processing, and analysis of texts. Documents and associated document- and collection-level metadata are easily loaded and stored as a corpus object, although most of quanteda's operations work on simple character objects as well. A corpus is designed to efficiently store all of the texts in a collection, as well as meta-data for documents and for the collection as a whole. This makes it easy to perform natural language processing on the texts in a corpus simply and quickly, such as tokenizing, stemming, or forming ngrams. quanteda's functions for tokenizing texts and forming multiple tokenized documents into a document-feature matrix are both extremely fast and extremely simple to use. quanteda can segment texts easily by words, paragraphs, sentences, or even user-supplied delimiters and tags.
Works nicely with UTF-8. Built on the text processing functions in the stringi package, which is in turn built on C++ implementation of the ICU libraries for Unicode text handling, quanteda pays special attention to fast and correct implementation of Unicode and the handling of text in any character set, following conversion internally to UTF-8.
Built for efficiency and speed. All of the functions in quanteda are built for maximum performance and scale while still being as R-based as possible. The package makes use of three efficient architectural elements: the stringi package for text processing, the Matrix package for sparse matrix objects, and the data.table package for indexing large documents efficiently. If you can fit it into memory, quanteda will handle it quickly. (And eventually, we will make it possible to process objects even larger than available memory.)
Super-fast conversion of texts into a document-feature matrix. quanteda is principally designed to allow users a fast and convenient method to go from a corpus of texts to a selected matrix of documents by features, after defining and selecting the documents and features. The package makes it easy to redefine documents, for instance by splitting them into sentences or paragraphs, or by tags, as well as to group them into larger documents by document variables, or to subset them based on logical conditions or combinations of document variables. A special variation of the "dfm", a feature co-occurrence matrix, is also implemented, for direct use with embedding and representational models such as text2vec.
Extensive feature selection capabilities. The package also implements common NLP feature selection functions, such as removing stopwords and stemming in numerous languages, selecting words found in dictionaries, treating words as equivalent based on a user-defined "thesaurus", and trimming and weighting features based on document frequency, feature frequency, and related measures such as tf-idf.
Qualitative exploratory tools. Easily search and save keywords in context, for instance, or identify keywords. Like all of quanteda's pattern matching functions, users have the option of simple "glob" expressions, regular expressions, or fixed pattern matches.
Dictionary-based analysis. quanteda allows fast and flexible implementation of dictionary methods, including the import and conversion of foreign dictionary formats such as those from Provalis's WordStat, the Linguistic Inquiry and Word Count (LIWC), Lexicoder, Yoshioder, and YAML.
Text analytic methods. Once constructed, a dfm can be easily analyzed using either quanteda's built-in tools for scaling document positions (for the "wordfish" and "Wordscores" models, direct use with the ca package for correspondence analysis), predictive models using Naive Bayes multinomial and Bernoulli classifiers, computing distance or similarity matrixes of features or documents, or computing readability or lexical diversity indexes.
In addition, quanteda a document-feature matrix is easily used with or converted for a number of other text analytic tools, such as:
topic models (including converters for direct use with the topicmodels, LDA, and stm packages);
machine learning through a variety of other packages that take matrix or matrix-like inputs.
Planned features. Coming soon to quanteda are:
Bootstrapping methods for texts that makes it easy to resample texts from pre-defined units, to facilitate computation of confidence intervals on textual statistics using techniques of non-parametric bootstrapping, but applied to the original texts as data.
Additional predictive and analytic methods by expanding the
textstat_andtextmodel_functions. Current textmodel types include correspondence analysis, "Wordscores", "Wordfish", and Naive Bayes; current textstat statistics are readability, lexical diversity, similarity, and distance.Expanded settings for all objects, that will propogate through downstream objects.
Object histories, that will propogate through downstream objects, to enhance analytic reproducibility and transparency.
How to Install
From CRAN: Use your GUI's R package installer, or execute:
install.packages("quanteda")From GitHub, using:
# devtools packaged required to install quanteda from Github devtools::install_github("kbenoit/quanteda")Because this compiles some C++ source code, you will need a compiler installed. If you are using a Windows platform, this means you will need also to install the Rtools software available from CRAN. If you are using OS X, you will need to to install XCode, available for free from the App Store, or if you prefer a lighter footprint set of tools, just the Xcode command line tools, using the command
xcode-select --installfrom the Terminal.Additional recommended packages:
The following packages work well with or extend quanteda and we recommend that you also install them:
readtext: An easy way to read text data into R, from almost any input format.
spacyr: NLP using the spaCy library, including part-of-speech tagging, entity recognition, and dependency parsing.
quantedaData: Additional textual data for use with quanteda.
devtools::install_github("kbenoit/quantedaData")LIWCalike: An R implementation of the Linguistic Inquiry and Word Count approach to text analysis.
devtools::install_github("kbenoit/LIWCalike")
Getting Started
See the package website, which includes the Getting Started Vignette.
Demonstration
library(quanteda)
## quanteda version 0.99.12
## Using 7 of 8 threads for parallel computing
##
## Attaching package: 'quanteda'
## The following object is masked from 'package:utils':
##
## View
# create a corpus from the immigration texts from UK party platforms
uk2010immigCorpus <-
corpus(data_char_ukimmig2010,
docvars = data.frame(party = names(data_char_ukimmig2010)),
metacorpus = list(notes = "Immigration-related sections of 2010 UK party manifestos"))
uk2010immigCorpus
## Corpus consisting of 9 documents and 1 docvar.
summary(uk2010immigCorpus)
## Corpus consisting of 9 documents:
##
## Text Types Tokens Sentences party
## BNP 1125 3280 88 BNP
## Coalition 142 260 4 Coalition
## Conservative 251 499 15 Conservative
## Greens 322 679 21 Greens
## Labour 298 683 29 Labour
## LibDem 251 483 14 LibDem
## PC 77 114 5 PC
## SNP 88 134 4 SNP
## UKIP 346 723 27 UKIP
##
## Source: /Users/kbenoit/Dropbox (Personal)/GitHub/quanteda/* on x86_64 by kbenoit
## Created: Fri Oct 6 09:35:32 2017
## Notes: Immigration-related sections of 2010 UK party manifestos
# key words in context for "deport", 3 words of context
kwic(uk2010immigCorpus, "deport", 3)
##
## [BNP, 157] The BNP will | deport | all foreigners convicted
## [BNP, 1946] . 2. | Deport | all illegal immigrants
## [BNP, 1952] immigrants We shall | deport | all illegal immigrants
## [BNP, 2585] Criminals We shall | deport | all criminal entrants
# create a dfm, removing stopwords
mydfm <- dfm(uk2010immigCorpus, remove = stopwords("english"), remove_punct = TRUE)
mydfm
## Document-feature matrix of: 9 documents, 1,547 features (83.8% sparse).
topfeatures(mydfm, 20) # 20 top words
## immigration british people asylum britain uk
## 66 37 35 29 28 27
## system population country new immigrants ensure
## 27 21 20 19 17 17
## shall citizenship social national bnp illegal
## 17 16 14 14 13 13
## work percent
## 13 12
# plot a word cloud
set.seed(100)
textplot_wordcloud(mydfm, min.freq = 6, random.order = FALSE,
rot.per = .25,
colors = RColorBrewer::brewer.pal(8,"Dark2"))Contributing
Contributions in the form of feedback, comments, code, and bug reports are most welcome. How to contribute:
Fork the source code, modify, and issue a pull request through the project GitHub page. See our Contributor Code of Conduct and the all-important quanteda Style Guide.
Issues, bug reports, and wish lists: File a GitHub issue.
Usage questions: Submit a question on the quanteda channel on StackOverflow.
Contact the maintainer by email.