corpus_chunk: Segment a corpus into chunks of a given size

In quanteda: Quantitative Analysis of Textual Data

Segment a corpus into chunks of a given size

Description

Segment a corpus into new documents of roughly equal sized text chunks, with the possibility of overlapping the chunks.

Usage

corpus_chunk(
  x,
  size,
  truncate = FALSE,
  use_docvars = TRUE,
  verbose = quanteda_options("verbose")
)

Arguments

x	tokens object whose token elements will be segmented into chunks
size	integer; the (approximate) token length of the chunks. See Details.
truncate	logical; if TRUE, truncate the text after size
use_docvars	if TRUE, repeat the docvar values for each chunk; if FALSE, drop the docvars in the chunked tokens
verbose	if TRUE print a summary of the object returned by the function: the number of documents and, depending on the object, the number of characters, tokens and types, or features and occurrences. The number of tokens includes paddings, but the number of features excludes them.

Details

The token length is estimated using stringi::stri_length(txt) / stringi::stri_count_boundaries(txt) to avoid needing to tokenize and rejoin the corpus from the tokens.

Note that when used for chunking texts prior to sending to large language models (LLMs) with limited input token lengths, size should typically be set to approximately 0.75-0.80 of the LLM's token limit. This is because tokenizers (such as LLaMA's SentencePiece Byte-Pair Encoding tokenizer) require more tokens than the linguistically defined grammatically-based tokenizer that is the quanteda default. Note also that because of the use of stringi::stri_count_boundaries(txt) to approximate token length (efficiently), the exact token length for chunking will be approximate.

See Also

tokens_chunk()

Examples

data_corpus_inaugural[1] |>
  corpus_chunk(size = 10)

quanteda documentation built on Oct. 21, 2026, 5:07 p.m.