Construct a corpus object

Creates a corpus object from available sources. The currently available sources are:

  • a character vector, consisting of one document per element; if the elements are named, these names will be used as document names.

  • a data.frame (or a tibble tbl_df), whose default document id is a variable identified by docid_field; the text of the document is a variable identified by textid_field; and other variables are imported as document-level meta-data. This matches the format of data.frames constructed by the the readtext package.

  • a kwic object constructed by kwic.

  • a tm VCorpus or SimpleCorpus class object, with the fixed metadata fields imported as docvars and corpus-level metadata imported as metacorpus information.

  • a corpus object.

corpus(x, ...)

# S3 method for corpus corpus(x, docnames = quanteda::docnames(x), docvars = quanteda::docvars(x), metacorpus = quanteda::metacorpus(x), compress = FALSE, ...)

# S3 method for character corpus(x, docnames = NULL, docvars = NULL, metacorpus = NULL, compress = FALSE, ...)

# S3 method for data.frame corpus(x, docid_field = "doc_id", text_field = "text", metacorpus = NULL, compress = FALSE, ...)

# S3 method for kwic corpus(x, split_context = TRUE, extract_keyword = TRUE, ...)

# S3 method for Corpus corpus(x, metacorpus = NULL, compress = FALSE, ...)


a valid corpus source object


not used directly


Names to be assigned to the texts. Defaults to the names of the character vector (if any); doc_id for a data.frame; the document names in a tm corpus; or a vector of user-supplied labels equal in length to the number of documents. If none of these are round, then "text1", "text2", etc. are assigned automatically.


a data.frame of document-level variables associated with each text


a named list containing additional (character) information to be added to the corpus as corpus-level metadata. Special fields recognized in the summary.corpus are:

  • source a description of the source of the texts, used for referencing;

  • citation information on how to cite the corpus; and

  • notes any additional information about who created the text, warnings, to do lists, etc.


logical; if TRUE, compress the texts in memory using gzip compression. This significantly reduces the size of the corpus in memory, but will slow down operations that require the texts to be extracted.


optional column index of a document identifier; defaults to "doc_id", but if this is not found, then will use the rownames of the data.frame; if the rownames are not set, it will use the default sequence based on (quanteda_options("base_docname").


the character name or numeric index of the source data.frame indicating the variable to be read in as text, which must be a character vector. All other variables in the data.frame will be imported as docvars. This argument is only used for data.frame objects (including those created by readtext).


logical; if TRUE, split each kwic row into two "documents", one for "pre" and one for "post", with this designation saved in a new docvar context and with the new number of documents therefore being twice the number of rows in the kwic.


logical; if TRUE, save the keyword matching pattern as a new docvar keyword


The texts and document variables of corpus objects can also be accessed using index notation. Indexing a corpus object as a vector will return its text, equivalent to texts(x). Note that this is not the same as subsetting the entire corpus -- this should be done using the subset method for a corpus.

Indexing a corpus using two indexes (integers or column names) will return the document variables, equivalent to docvars(x). It is also possible to access, create, or replace docvars using list notation, e.g.

myCorpus[["newSerialDocvar"]] <- paste0("tag", 1:ndoc(myCorpus)).

For details, see corpus-class.


A corpus-class class object containing the original texts, document-level variables, document-level metadata, corpus-level metadata, and default settings for subsequent processing of the corpus.

A warning on accessing corpus elements

A corpus currently consists of an S3 specially classed list of elements, but you should not access these elements directly. Use the extractor and replacement functions instead, or else your code is not only going to be uglier, but also likely to break should the internal structure of a corpus object change (as it inevitably will as we continue to develop the package, including moving corpus objects to the S4 class system).

See Also

corpus-class, docvars, metadoc, metacorpus, settings, texts, ndoc, docnames

  • corpus
  • corpus.corpus
  • corpus.character
  • corpus.kwic
  • corpus.Corpus
# create a corpus from texts

# create a corpus from texts and assign meta-data and document variables
               docvars = data.frame(party = names(data_char_ukimmig2010))), 5)


# import a tm VCorpus
if (requireNamespace("tm", quietly = TRUE)) {
    data(crude, package = "tm")    # load in a tm example VCorpus
    mytmCorpus <- corpus(crude)
    summary(mytmCorpus, showmeta=TRUE)

    data(acq, package = "tm")
    summary(corpus(acq), 5, showmeta=TRUE)

    tmCorp <- tm::VCorpus(tm::VectorSource(data_char_ukimmig2010))
    quantCorp <- corpus(tmCorp)

# construct a corpus from a data.frame
dat <- data.frame(letter_factor = factor(rep(letters[1:3], each = 2)),
                  some_ints = 1L:6L,
                  some_text = paste0("This is text number ", 1:6, "."),
                  stringsAsFactors = FALSE,
                  row.names = paste0("fromDf_", 1:6))
summary(corpus(dat, text_field = "some_text",
               metacorpus = list(source = "From a data.frame called mydf.")))

# construct a corpus from a kwic object
mykwic <- kwic(data_corpus_inaugural, "southern")
# from a kwic
kw <- kwic(data_char_sampletext, "econom*")
summary(corpus(kw, split_context = FALSE))
texts(corpus(kw, split_context = FALSE))

# }
Documentation reproduced from package quanteda, version 1.4.1, License: GPL-3

Community examples

Looks like there are no examples yet.